US20230349575A1 - Exhaust demand control system and methods - Google Patents

Abstract

Methods and apparatus for an exhaust demand control system for measuring one or more contaminants at one or more exhaust locations within one or a plurality of exhaust ducts or plenums served by an exhaust fan system. Example systems and methods can include sensing the one or more contaminants within the one or more exhaust duct locations using a multipoint air sampling system having one or more sensors and comparing contaminant concentration measurements from the one or more of said exhaust duct or plenum locations against an action level to create a fan setback signal.

Description

CROSS REFERENCE TO RELATED APPLICATIONS
• The present application is a continuation of U.S. Pat. Application No. 17/822,474, filed on Aug. 26, 2022, which is a Divisional of U.S. Pat. Application No. 16/141,109, filed on Sep. 25, 2018 now issued as U.S. Pat. No. 11,460,203 on Oct. 4, 2022, which claims the benefit of U.S. Provisional Pat. Application No. 62/581,877, filed on Nov. 6, 2017, all of which are incorporated herein by reference.
FIELD OF THE INVENTION
• The present invention relates generally to the energy efficient operation of an exhaust fan system and, more particularly, to systems and methods used to monitor the presence of contaminates in exhaust air, and/or reducing risk when controlling exhaust fans in order to optimize exhaust fan operation in a safe manner. Some embodiments may be well suited to optimize energy use with high plume exhaust fan systems.
BACKGROUND
• There are a broad range of facilities that have ventilation systems which are designed to safely support the use of chemical and biological compounds which have exposure limits at which occupant health, comfort, and productivity can be affected. This includes, but is not limited to, facilities designed for research, experimentation, production operations, testing, health care, animal and pharmaceutical research, and other applications.
• Generally, ventilation energy use in these types of facilities is significant as they are often designed as what’s known in the art as “single pass” ventilation systems. In these types of systems, air supplied to a room or critical location designed to handle contaminant use cannot be recirculated but must be fully exhausted from the facility after it is drawn from the critical location by the ventilation system. ASHRAE Standard 62.1-2016, which is incorporated herein by reference, defines ventilation practices for building locations based upon four categories of risk, which determines whether or not air can be recirculated to a zone. As is known in the art, only air from spaces described by ASHRAE 62.1-2016 asClass 1 environments can be recirculated to other locations in a building.Class 1 environments are generally categorized as locations, such as office environments, that have air with low contaminant concentration, low sensory-irritation intensity, and inoffensive odor. According to the standard,Class 2 air includes air that is not necessarily harmful but that is inappropriate for transfer or recirculation to spaces used for different purposes.Class 2 air may be recirculated within the space of origin, but not toClass 1 space. As is known in the art,Class 3 air is that which may have significant contaminant concentration, significant sensory-irritation intensity, or offensive odor.Class 3 air may be recirculated within the space of origin, but not to any other space. Air from most lab environments is considered asClass 3 air and therefore needs to be completely exhausted from the building.Class 4 air is considered as potentially the most harmful or objectionable. As an example, the air from a fume hood would be considered asClass 4. According to ASHRAE 62.1-2016,Class 4 air shall not be recirculated or transferred to any space and not be recirculated within the space of origin.
• Whether or not air is exhausted from a building or is partially or wholly recirculated, influences the amount of heating and cooling energy as well as fan energy that must be provided. Generally, when air is allowed to be recirculated within a building the heating and cooling energy as well as the fan energy use will be substantially lower.
• FIG.1 is a simplified illustration of a prior art ventilation system which incorporates recirculated air from 5 zones. The ventilation system shown inFIG.1 includes a supply fan and a return fan. Those familiar with the art of ventilation systems will recognize that there are a number of different fan configurations used in recirculating air systems (also known as mixed air systems). The supply fan delivers supply air to each zone via a common supply plenum. A portion of the total supply airflow is delivered to each zone via zone boxes (ZB-1, ZB-2, ZB-3, ZB-4, and ZB-5) which may include any method known to the art to adjust airflow. In actual application, the supply fan may serve any number of zones and airflow levels delivered to each zone will vary based on space cooling, heating and outside air ventilation requirements. With a return air system such as that ofFIG.1, the return fan draws air form each zone and a portion of that air is recirculated back to the supply fan inlet via the return damper. In a system such as this, when more outside air must be brought into the building, the return air flow will be proportionally decreased. Thus, there is an inverse relationship between outside air and return air flow levels. However, there is generally a direct relationship between outside air intake levels and the building exhaust airflow with a mixed air system. With systems like this, the building exhaust will be balanced against the outside air intake to ensure good building pressurization. Often times, the building exhaust will be set to a slightly lower flow rate than the outside air intake to ensure that the building is positively pressurized. The air returned by a system like this should only be fromClass 1 spaces (such as general office space), as air form higher risk zones (Class 2, 3, and 4) should not be recirculated in accordance with ASHRAE guidelines. However, ASHRAE does permit air supplied from a mixed air system to also supply air to more critical spaces, such asClass 2, 3, and 4 spaces. For example, the supply air inFIG.1 could be used as ventilation for a laboratory space. It would not be permissible however to recirculate the air from that lab.
• FIG.2 is a generalized illustration of a typical prior art ventilation system used with labs and other critical spaces. For simplicity, excludingcorridor217, only four zones (1B, 2B, 3B, and 4B) are depicted inFIG.2, although those experienced in the art of lab ventilation will appreciate that such systems may serve many more locations than that shown and in some cases less. Note that devices (201 and219) called venturi valves (also called “valves”) are shown instead of dampers to control supply and exhaust air because of their common use in labs; however, this description applies to any flow control devices used in the art. Systems such as this are commonly referred to as 100% outside air systems because the supply fan (202) in this case only draws air from outside the building and does not incorporate mixed air. As previously stated, ASHRAE does permit the supply of mixed air to critical spaces. However, if most or all of the zones, such aszones 1B, 2B, 3B and 4B inFIG.2 are classified asClass 2, 3, or 4 spaces, all of the air exhausted from these spaces via204,205,207,208, and211) has to be conveyed by theexhaust fan203 and discharged from the building and none of this air can be recirculated or distributed to the other locations. It’s actually common practice to even incorporate air fromClass 1 spaces with lab exhaust to avoid the cost of having to install and operate two different fan systems.FIG.2 also illustrates another common aspect of many conventional lab systems which incorporate office space, such aszone 3B. Inzone 3B ofFIG.2, onlysupply air210 is provided to the office space and this results inairflow215 betweenzone 3B and 2B that is equivalent to that supplied to the office location. This practice helps to reduce the number ofexhaust valves201 or dampers that need to be installed, which reduces cost and complexity. Given thatzone 3B has been defined as office space, it is essentially aClass 1 environment and therefore the air flowing fromzone 3B tozone 2B can be considered to be relatively free of contaminates and can help to dilute any contaminants that may be present inzone 2B.
• In addition, each of thezones 1B, 2B, and 4B inFIG.2 are connected to acommon corridor217 and the corridor would also normally be considered to be aClass 1 environment. Therefore, the air which flows from the common corridor into each space also provides some dilution to contaminants which may be present in the critical lab zones. The flow of air from one contiguous location to another is often referred to in the art as “room offset”. Typically, room offset (often called volumetric offset) is a design parameter used to establish the pressurization of one zone in relation to that of another. The volumetric offset is often calculated as the difference between the total air that is mechanically supplied to a space subtracted from the total air that is mechanically exhausted from the space. Therefore, as an example, if500 cubic feet per minute (CFM) is supplied (206) tozone 1B and200 CFM is exhausted through the fume hood (205) in that space and400 CFM is removed through the general exhaust (204) in that space, the volumetric offset inzone 1B (213) will be -100 CFM. Therefore, in this example,zone 1B will be negatively pressurized with respect to the corridor and air (213) will flow at a rate of100 CFM from the corridor tozone 1B. Labs are often configured to operate with a negative offset to prevent contaminants that might be released from a chemical spill within the lab from spreading to other locations.
• As is known in the art, fume hoods (such as205 and208) are often incorporated within labs to allow lab personnel to safely conduct work, such as experimentation, with compounds that may pose a health hazard or present an objectionable odor or sensory irritation. Fume hoods are normally setup to draw at least some minimum level of airflow from a given lab space in order to ensure that any spill of contaminants, such as chemical compounds, within the hood cannot travel into the lab or expose personnel working at the hood. It is common practice to vary the amount of airflow drawn into the hood as a function of the hood sash opening, using variable volume controls. U.S Pat. No. 4,706,553, incorporated herein, is an example of the intricate operation of a variable volume fume hood controller. U.S. Pat. No. 4,893,551 and U.S. Pat. No. 6,137,403, which are also incorporated herein, are examples of fume hood sash sensing approaches used to vary fume hood air flow rates with sash opening. Any number of fume hoods may be present within a laboratory and, the amount of airflow exhausted through a fume hood will generally vary with the size of the hood and its sash position. Fume hoods are generally considered asClass 4 environments, as described by ASHRAE 62.1-2016.
• Although fume hoods will at times exhaust airborne contaminants into an exhaust plenum (such as220) at concentrations that would be unhealthy or objectionable for lab occupants to breath, most of the time the air flowing from a fume hood into the exhaust system will be relatively clean. This in part is due to the fact that most fume hoods are not under continuous use by lab personnel. Nevertheless, there will normally be some continuous amount of airflow through most fume hoods (such as205 and208), based upon guidelines provided by ANSI Z9.5-2012.
• The general exhaust (such as204,207, and211) is provided in labs in order to ensure the desired lab pressurization and to provide an exhaust path for contaminant spills that may take place in the lab. As with fume hood controls, the air flow controls for each lab space may operate either as what’s known in the art as constant volume systems or variable volume systems. When the flow controls include exhaust and supply airflow devices such as201 and219 that have fixed flow settings, those experienced in the art of lab ventilation would recognize that the airflow system would be referred to as a constant volume system. Constant volume systems are generally less energy efficient than variable volume systems, as they tend to apply more ventilation to a given lab space than is necessary because they must be fixed to deliver the worst-case ventilation needs.
• Those experienced with laboratory ventilation controls will appreciate that most ventilation control strategies are designed to satisfy the relationship ofEquation 1 below.
• $\begin{array}{cc}\begin{array}{l}LabSupplyCFM=\\ GeneralExhaustCFM+FumeHoodCFM+VOlumeOffset\end{array}& \text{­­­(Equation 1)}\end{array}$
• There are many ways to satisfy the ventilation balance ofEquation 1 and in the art, the control strategy varies based on the manufacturer of the ventilation control system and the preferences of the specifying engineer. In some cases, such as with room and lab environments that are tightly sealed the volume offset may be actively varied in order to control the pressure of the room or lab space based upon a pre-determined pressure setpoint. U.S. Pat. No. 5,385,505 A, incorporated herein, describes a pressure maintenance system for substantially sealed spaces.
• Unless there are chemical spills or the lab chemical handling protocol and hygiene is poor, the air quality in labs is generally quite good. This is due to the fact that ventilation rates in labs are generally much higher than that of less critical environments, such as office spaces andother Class 1 spaces. As is known in the art, a figure of merit which is used to describe ventilation levels is “air change rate”, which is often measured as air changes per hour or ACH. This is a measure of the number of times per hour the air in a room is fully replaced or exchanged with fresh new air. Over recent years, there has been a trend in the Heating, Ventilation, and Air Conditioning (HVAC) community to decrease air change rates within laboratory environments and other critical spaces. For example, in the 1990s it was quite common to specify air change rates of 12 ACH or higher in labs. Historically, it was also common to specify 18 ACH or more in animal facilities or vivariums. One influence on this tendency is that ANSI Z9.5-2012 states that “... air changes per hour is not the appropriate concept for designing contaminant control systems.” A guideline for animal facilities that is frequently referenced is the Guide for the Care and Use of Laboratory Animals by the Institute for Laboratory Animal Research (ILAR) which states that a “Provision of 10 to 15 fresh air changes is an acceptable guideline” but that the “use of such a broad guideline (for animal rooms) might over-ventilate a macro-environment containing few animals...”
• Another factor which has resulted in reduced air changes in critical environments, such as labs, is that the thermal loads in many of today’s labs are quite low in comparison to what they were 10 to 15 years ago. One influence on this is the use of higher efficiency lighting in labs, such as LED-based lighting technology. Also, personal computers that are used in labs now have energy efficient LCD screens, which use only a fraction of the power of the older CRT-based monitors. More energy efficient technologies such as these have significantly reduced the amount of added heat given off by equipment in the lab space. This results in a lower overall wattage per square foot of equipment related heat gain in the lab space, thereby reducing the cooling requirements for these spaces. For example, it used to be common for labs to be designed with a thermal cooling load of 10 watts per square foot or higher. Now, most labs operate at 3 watts per square foot or less. Lab supply air flow rates are commonly used to handle the labs cooling load. As the lab cooling load is reduced, the supply air flow requirements also reduce.
• Today, it is quite common for engineers to specify 6 ACH for occupied hours in labs and as little as 2 ACH during unoccupied hours. With the application of an active monitoring system used to sense for lab contaminants, it has also become common to specify 4 ACH during occupied hours in labs. U.S. Patent No. 6,425,297, which is incorporated herein by reference, describes a system that can be used for such room level monitoring purposes. Also, in animal rooms today engineers are specifying air change rates of 10 ACH or less.
• In a 2017 ASHRAE Technology Award Case Study [ASHRAE Journal, July 2017], incorporated by reference herein, Crowley describes substantial flow reduction energy conservation measures that include the application of active chilled beams. Those experienced in the art recognize that chilled beams or radiant cooling coils utilize chilled water and not air to remove heat from the room. Chilled beams are radiator like coils through which chilled water is flowed in order to provide a chilled surface that is usually located in the room’s ceiling. Because of the relatively cool surface, convective airflow is established in the space as relatively warm room air rises to the ceiling-mounted coil. The cool air that emanates from the coil has a downward flow that acts as a mechanism to convey cool air to room occupants. The use of chilled beam technology has become prevalent in lab and non-lab buildings alike and helps to reduce the amount of supply air needed in a space for cooling purposes.
• Even though there has been a trend for reductions in lab ventilation rates, air quality in most labs and other critical environments tends to be very good with few airborne contaminants in the lab space the majority of the time. In an ASHRAE Journal article “Demand-Based Control of Lab Air Change Rates” [Sharp, ASHRAE Journal, February 2010], incorporated herein, Sharp presents data representing a large number of labs taken over 1.5 million hours of operation. The data shows that the labs of this study were relatively free from contaminants more than 99% of the time.
• Exhaust fan203 is used to convey the air through each exhaust control device (201) and expel the air from the building. In its operation,exhaust fan203 must have the capacity to draw the required airflow through each of201. The combined exhaust of every zone (1B, 2B, 4B) is herein referred to as the “Total System Exhaust” (221), which is the total amount of air that needs to be drawn from the building locations served byfan203. As is known in the art, in practice the total system exhaust221 may be composed of exhaust air from any number of zones and may differ considerably from the example ofFIG.2. Also, because of variable air volume (VAV) zones the total system exhaust221 may vary by a considerable amount, due to VAV fume hood use and variations to the ventilation rates in each zone that may result from flow levels required for temperature control, as well as a number of variables that may prompt dynamic changes to each zone’s minimum ventilation rate. For example, it is common practice to operate a lab zone at one air change rate (for example, 2 ACH) during unoccupied hours and another air change rate (for example 6 ACH) during occupied hours. In all cases, theexhaust fan203 must provide enough negative pressure to the exhaust plenum (220) and acrossexhaust devices201 to ensure that the desired exhaust air flow is maintained at each zone under all conditions.
• Another important function ofexhaust fan203 is to expel the total system exhaust221 at a sufficient rate to ensure that any possible contaminants within221 are properly dispersed into the outdoor atmosphere so that these contaminants will not be entrained into locations within the building envelope (such as outside air intakes for example) or expose locations of neighboring buildings. In older ventilation system designs (those implemented before the 1990′s for example) it was quite common to ensure good dispersion of contaminants into the atmosphere by constructing a very tall exhaust stack on the building to whichfan203 would connect. These stacks would often reach heights of 30 to 40 feet, or higher, above the roof of the building, in order to ensure good dispersion performance of potential contaminants. The problem with structures such as this is that they are an esthetically unpleasing component of the building’s architecture, because of their size and the guy wires and other mechanical framework needed to support these structures. Also, because of the guy wires needed to support these stacks, it can be difficult to erect multiple stacks such as this on roofs with limited space.
• A popular alternative to the use of the aforementioned large exhaust stack has been high plume exhaust fans or high plume dilution fans, herein referred to as high plume fans. High plume fans provide a way to create an effective stack height that is many times the actual physical height of the fan stack. This is accomplished using a nozzle design that is integrated with the outlet of the fan which increases the discharge or exit velocity of the exhaust flow. High plume fans also may incorporate a bypass airflow control element which introduces quantities of outdoor air with the total system exhaust flow (221) to ensure that a target fan exit velocity is maintained by the fan system at all times. The bypass air will vary inversely with the total system exhaust flow, which results in a constant total airflow at the fan’s outlet. Using a high plume fan enables the fan to be relatively concealed, as the physical height of this type of exhaust stack system is typically only 10 to 15 feet in height.
• U.S. Pat. No. 4,806,076, which is incorporated herein by reference describes an early high plume fan design, which resembles many of the fan systems used today. Examples of commercially available high plume fan products include but are not limited to: Tri-Stack^® by Strobic Air Technologies, Axijet^® by M.K. Plastics Corporation, and Vektor^® series fans by Greenheck Inc.; brochures for each are incorporated herein. These are just examples of products which are commonly seen in use; those who are experienced in the art of ventilation systems will appreciate that there are a wide range of high plume fans made by numerous manufacturers.
• FIG.3A further illustrates the operation of a typical prior art highplume fan system300A, that may be used asexhaust fan203. Thesystem300A includes three discrete fans (309,310, and311) which have been manifolded together into onecommon plenum307. Note that any number of fans can be utilized in a high plume fan assembly but that three fans are shown inFIG.3A for illustrative purposes.
• A feature that is common with most types of high plume fans is the nozzle andwind band assembly313. As is known to those familiar with ventilation systems, a wind band provides protection to the exit nozzle from wind and weather conditions and it also will entrain addedairflow318.Airflow318 is also often referred to as dilution air since it contributes to the overall dilution performance of theexhaust fan203.
• It is also common practice to incorporate a backup fan in systems such as300A, given the critical nature of the operation of these systems. When a backup fan is incorporated, as one active fan fails,fan system300A operation will be maintained by activating the backup fan and taking the failed fan offline. For example,fan300A could be configured with one of the three fans serving as backup. One of the features incorporated with most fan systems such as300A is that they often incorporate shut offdampers308. The purpose for the shutoff damper assigned to each fan (309,310, and311) is to provide a way to take a fan offline or to provide a way to enable a fan to serve as a backup. For example, iffan309 serves as backup tosystem300A, itsshutoff damper308 will be closed so that no air is allowed to flow throughfan309 when it is not running. During this time, the shutoff dampers to the fans which are running (310 and311) will be open so that air can flow fromplenum307 into310 and311 and then the312 discharge air. It is also common practice for anexhaust fan assembly300A to periodically change which fans are active and which fan serves as the backup fan. For example, during one period oftime fan309 may serve as the backup, withfans310 and311 serving as the active fans. During another period of time,fan310 may serve as backup withfans309 and311 serving as the backup; and so on. This strategy of rotating the backup fan is known in the art as a “lead-lag” sequence, and it serves as a way to ensure that each fan ages in a uniform manner. The lead-lag term refers to the method used to determine which fan will be shut off to serve as backup, as the backup fan is made active. The strategy designates the fan which has been on for the longest period of time as the one which will next become the backup fan. This rotation or change to which fan may serve as backup normally happens every few days and is usually a parameter that is programmed into the control logic for300A. One practical issue that sometimes results in cold climates with the lead-lag function is that theshutoff damper308 on the backup fan can become frozen due to ice buildup. When this occurs, as thefan system300A initiates the lead-lag function it can result in a compromise in theflow delivery312, due to the fact that no airflow will result from the fan being activated. As a result, in colder climates where there may be ice and snow buildup, it is not uncommon to eliminate the lead-lag function during the winter months. However, that does detract from the uniformity with which the fan systems will age, resulting in early failures in some fan components such as bearings for example.
• FIG.3A also illustrates the connection of four duct risers (302,303,304, and305). With exhaust systems it is quite common to have a number of duct risers which run vertically through the building to connect various locations. For example, it’s common to have a separate vertical exhaust riser per floor or ones that interconnect different wings of a building. It would be apparent to those skilled in the art of ventilation systems that any number of risers could connect toplenum307. In some cases, for example, there will be one exhaust riser that may connect one or more manifolded ducts on one or a plurality of floors in the building. However, the configuration of a plurality of risers as shown inFIG.3A is quite common. Also, it is common practice to interconnect any number of exhaust risers to acommon plenum307 in the mechanical space or penthouse that is located just below theroof306.
• In a typicalsystem bypass air301 is mixed with the total system exhaust (221) which for300A would be the combination ofexhausts314,315,316, and317.Bypass flow301 is adjusted until the desired exit velocity of312 is achieved. In many applications today, it is common to set this exit velocity to 3000 feet per minute. An exit velocity of 3000 feet per minute (fpm) is often specified based on guidance from ANSI Z9.5-2012. However, the ANSI standard specifically states that 3000 fpm “is required unless it can be demonstrated that a specific design meets the dilution criteria necessary to reduce the concentration of hazardous materials in the exhaust to safe levels...”
• There are several common ways in which thefan system300A will be controlled. Usually, the active fans (the ones which are not selected to be in backup or standby mode) will be set to operate at a fixed speed so that each active fan will be able to deliver the desired exit velocity (for example 3000 fpm). In most cases, the control of thebypass air301 will vary inversely with the Total System Exhaust (221). The most common way to controlbypass301 is by controlling a modulating damper to vary301 in order to maintain a fixed static pressure setpoint withinplenum307. For example, thebypass301 may be varied to control the static pressure to -4 inches of water column (inches H20), but setpoints in different applications may vary considerably.FIG.3A shows a bypass inlet anddamper assembly301 on each side ofexhaust fan system300A; however, in some cases there will be only onemain inlet301. In other cases, there may be a plurality ofbypass inlets301. The total airflow exiting an exhaust fan system is a function of fan speed. In some applications, fan speed is established by what’s known in the art as sheave settings on the fan assembly. In other cases where the total system exhaust221 may vary considerably, the exhaust fan speed will be controlled by way of a setpoint to one or more variable speed drives (VFD’s). VFD use is quite common with high plume exhaust fan systems because it provides an efficient way to control the power to the exhaust fan motors. Typically, each fan (309,310,311) will have a dedicated VFD.
• In many high plume fan applications, it is quite common for thebypass airflow301 drawn into theexhaust fan system300A to be a substantial portion of theoverall outlet airflow312. This is especially the case when measures have been taken to reduce the total system exhaust CFM (221). For example, as previously described, it has become quite common to operate critical spaces such as labs at 6 ACH during occupied hours and 2 ACH during unoccupied hours. There has been an increasing trend to implement such settings as an energy conservation measure (ECM) in existing labs which may have previously been operating at much higher minimum air change rates (for example12 ACH or higher). In buildings with high plume exhaust fan systems such as300A, most of the energy savings that results from lab flow reduction ECM’s is associated with supply airflow energy savings, as the reduction in airflow fromsupply fan202 results in lower supply fan energy use and there is also significant heating and cooling energy savings due to the reduction in the amount of outside air that needs to be brought into the building. On the exhaust side, however, as flows are reduced in the lab only the total system exhaust CFM (221) is reduced and not necessarily the outlet airflow of theexhaust fan312. With a system such as300A, as the total system exhaust CFM (221) is lowered thebypass air301 will proportionally be increased in order to maintain aconstant outlet flow312. As a result, there may be no energy savings realized in the operation ofexhaust fan system300B with the flow reduction ECM. In practice, there will usually be some exhaust fan energy savings with300A as a result of a lab flow reduction ECM if theexhaust fans309,310,311 are staged or as a result of reduced peak exhaust flows. Fan staging provides a way to save energy by reducing the number of active fans from300A when the total system exhaust CFM (221) is at a level where few fans can safely handle that flow rate221 while also operating at the desired outlet velocity (for example 3000 fpm). One of the problems with active staging of fans is that, similar to Lead-Lag operation, during winter months reliability issues can be encountered when attempting to turn fans on and off, due to the shut offdamper308 icing up. Therefore, in many of the colder climates in the world fan staging is not implemented.
• Table 1 below illustrates the operation of anexhaust fan system300A as the lab ventilation rate in an example building are reduced. Table 1 assumes a scenario where a building initially had spaces which on average operated at 12 ACH and, through a flow reduction ECM, now operate at 6 ACH on average. This building could have spaces that resemble those illustrated inFIG.2. For this example, however, the total system exhaust values shown in Table 1 are more representative of a larger facility, which is also more typical. As one can see from Table 1, when the average air change rate is 12 ACH in each location, the total system exhaust is 36,000 CFM. In that state, 16,000 CFM is brought into theplenum307 asbypass air301. Assuming that two of the three fans are active,52,000 CFM of exhaustfan total CFM312 is required in order to establish an exit velocity of 3,000 fpm at each fan. This is based on the cross-sectional area of the outlet nozzle on fan in300A. Those experienced in the art of ventilation systems will appreciate that actual fan geometries along with the room flow rates will vary considerably in practice and that the values here are specific only to this example. As the room air change rates are reduced to 6 ACH, the total system exhaust CFM will reduce to18,000 CFM as shown in Table 1. Table 1 considers two prior art scenarios of how theexhaust fans300A may be adjusted as the room air change rates are reduced to 6 ACH. This includes a non-staged fan strategy and strategy where the fans are staged. Again, in this example we assume that a maximum of only two fans of300A are running to handle the total system exhaust CFM. In the non-staged scenario, as the total system exhaust CFM is reduced to 18,000 CFM (due to ventilation rates being reduced to 6 ACH) two fans will continue to operate and therefore the exhaust fan total CFM will continue to be 52,000 CFM. This means that 34,000 CFM ofbypass air301 needs to be added to the total system exhaust CFM221 in order to maintain an exit velocity of 3000 fpm. With the staged fan scenario, because one fan will have enough capacity to handle the lower total system exhaust CFM221, one of the two active fans in300 will be shut off. In this example, two fans deliver 52,000 CFM, so one fan will deliver 26,000 CFM. In this scenario, 8,000 CFM ofbypass air301 must be combined with the 18,000 CFM of total system exhaust in order to deliver 26,000 CFM of exhaust fan total CFM, which is required to maintain an exit velocity of 3,000 fpm. Table 1 illustrates an aspect of thesesystems300A in that the bypass CFM values301 will often be a large percentage of the exhaust fan total CFM.
• TABLE 1

  Exhaust Fan Airflow with Ventilation Reduction ECM (Prior-Art)
  Average Room ACH	Total System Exhaust CFM	Exhaust Fan Total CFM	Bypass Air CFM
  12	36,000	52,000	16,000
  6 (non-staged)	18,000	52,000	34,000
  6 (staged)	18,000	26,000	8,000

• Bypass airflow301 values can be quite large, even when fans are staged. Because of this, high plume exhaust fan systems utilize a lot of energy and can be quite expensive to operate. In the example of a non-staged fan shown in Table 1, thebypass air301 represent 65% of the exhaust fan total CFM. Even for an efficiently operating fan, the amount of power required per CFM could easily be .7 Watts/CFM. At this rate it could require over 208,000 kilowatt hours (kWh) of electricity on an annual basis, just to operate thebypass air301 portion of thefan system300A. If for example the cost per kWh is $0.11 per kWh, this translates to over $22,880 just to run the bypass air, annually.
• Although a fan exit velocity of 3000 fpm will often be specified, in some cases even higher fan exit velocities will be specified because of a number of reasons that include but are not limited to: anticipated characteristics of the exhaust dispersion plume due to ambient wind speed and direction, physical structures (such as other buildings for example) which are in proximity to theexhaust fan system300A, and the unique dilution requirements due to usage quantities of the chemical inventory. This will further increase energy use as, to achieve higher exit velocitiesmore bypass air301 will generally be required.
• Another factor associated with theexhaust fan system300A performance is that, as the exit velocity offan300A is increased, acoustical noise within the human audible range can become a factor with these systems. Many of the world’s research facilities which may incorporate labs and high plume exhaust fans are located in metropolitan areas, including inner-city locations where strict noise threshold limitations may be implemented due to close building proximities. Runningfan systems300A at higher than necessary exit velocities can also result in higher radiated noise levels or sound pressure levels which will add to overall sound pressure levels emitted from the building, which can become a factor in meeting local noise regulations.
• Although theexhaust fan system300A represents a configuration of thegeneral fan system203, those familiar with the art of ventilation controls will recognize that a wide range of exhaust fan system configurations exist which do not utilize high plume fans. High plume fan use has become very popular, but many systems exist and continue to be specified which include fans other than the high plume style.FIG.3B illustrates an example of this, where a fan323 is connected through anoptional duct322 toplenum307, to exhaustair314,315,316, and317 which comingles inplenum307. The fan323 discharges intostack321 and this discharged air exits throughnozzle320. Moisture which may accumulate instack321 due to rain or condensation will drain throughdrain element324. Those familiar with the art of ventilation controls will realize that the fan configuration of300B is just one example of a wide range of exhaust fan configurations which is based on an approach which does not incorporate one or more high plume fans. For example, a common configuration includes but is not limited to centrifugal blowers that do not have an integratedwind band313. The fan element323 may also be an axial fan, a forward inclined fan, a reverse inclined fan, or any of a broad range of fan types that are known to those experienced in the art of ventilation controls. Thefan implementation300B may also incorporate a bypass air element in a manner that is like301 that’s illustrated in300A. Similar tosystem300A, which incorporates one or a plurality of fans (309,310,311),system300B may incorporate one or a plurality of fans323. Usually, when there are more than one fans323, each additional fan323 will also be configured with adedicated stack321 andnozzle320. As is known in the art, there are some conditions where more than one fan323 will be connected to acommon stack321 andnozzle312; this approach can be used to boost the exit velocity ofdischarge air312 which can be beneficial to improving the plume height and dispersive properties of thesystem300B. As is the case with the fan configuration ofsystem300A, depending on the requirements of the application,system300B may also be configured with any combination of numbers of risers and numbers of fans. In some cases, such as when there is only one riser (for example riser302),plenum307 may be reduced in size so that it is primarily the size ofduct322. Assuming there is abypass element301, this will often be implemented in a manner that is similar to that described forsystem300A.
• Methods of reducing high plume exhaust fan energy use has been a topic which has received a lot of attention by the HVAC engineering community over recent years. One approach to lowering high plume fan energy use that has been tried has been to incorporate active environmental sensing of contaminants within the total system exhaust221. This approach, herein referred to as exhaust demand control (also referred to in the art as “exhaust fan control”), has the objective of operating theexhaust fan system300A at two different potential exit velocities based on whether contaminants are detected in the exhaust stream221 or not. If contaminants are detected, then thefan system300A would be commanded to operate at a higher exit velocity (such as 3000 fpm or higher). If on the other hand the exhaust stream221 is determined to be relatively clean, thefan system300A would be commanded to operate at a lower exit velocity, such as a value as low as 750 fpm or another suitable exit velocity. Exhaust demand control may be applied to non-high plume fan systems as well, such assystem300B. Serious issues with contaminant sensing, as described further below, have prevented exhaust demand control from working effectively.
• In an ASHRAE Journal article “Saving Energy in Lab Exhaust Systems” [Carter et al., ASHRAE Journal, June 2011], incorporated herein, methods are presented to vary exhaust exit velocities based on whether the exhaust stream is relatively free of contaminants using a chemical monitoring system. Wind velocity measurement is also reviewed as an added approach. Wind velocity has an inverse influence on the effective stack height of asystem300A or300B which reduces the dispersion of contaminants as windspeed is increased. The concept is to therefore monitor windspeed (using an anemometer for example) and to save fan energy by reducing exhaust exit velocities during non-windy times. Windspeed monitoring by itself however yields only a very limited savings as exit velocities of 3000 fpm or more may still be required when exhaust air contains contaminants.
• FIG.4 illustrates two prior art approaches which have been implemented to detect contaminants in the total system exhaust221. One approach involves monitoring the total system exhaust with one or morediscrete sensors401, which are disposed within thecommon plenum307 to which the air from eachexhaust riser314,315,316, and317 will comingle. In this application,sensor401 communicates its reading to either the fan controls or the building automation system (BAS) that communicates with the fan controls. Logic may be setup either within thesensor401 electronics or within the BAS or fan controls to determine when thefan system 300A/300B can operate at a lower exit velocity or when it must operate at a pre-determined higher velocity.
• The sensor that is typically used for401 is known in the art as a photoionization detector or PID. PIDs are used extensively for a variety of environmental health and safety applications because of their ability to detect hundreds of different compounds and especially volatile organic compounds (VOCs). A PID can also detect a limited number of inorganic compounds as well. Volatile organic compounds are of special interest to applications such as400 because a large percentage of the chemical inventory that is used in labs and other critical environments which require the most amount of dilution or dispersion from the exhaust fan system are VOCs. U.S. Pat. No. 6,646,444, which is incorporated herein, describes an exemplary PID used in systems such as multiplexed air sampling systems.
• One characteristic of a photoionization detector is that it is that it is able to provide a signal that is simultaneously responsive to multiple compounds. This is sometimes referred to as a “broadband” sensing characteristic. Other types of broadband sensors include but are not limited to metal oxide semiconductor (MOS) sensors, flame ionization detectors, and total organic compound (TOC) infrared sensors. With a PID, the photoionization occurs as a molecule absorbs a photon of energy at a sufficient level to release an electron to create a positive ion. This takes place when the ionization potential of the molecule in electron volts (eV) is less than the energy of the photon. A PID uses a specialized ultraviolet lamp as its photonic source. It is common to use PIDs with lamps which operate at 10.6 eV, as these lamps tend to be reasonably durable for detecting compounds in most occupant environments while also providing a broad detection range. As a compound is ionized by the lamp, electron flow is measured by a detector electrode, and this current is proportional to the concentration of the gas that has been ionized. Different compounds can be ionized at a given time, allowing the sensor to be responsive to concentrations of multiple compounds. A PID is also a very sensitive device which, when used in relatively clean environments, can reliably detect many compounds at concentrations of a few tens of parts per billion.
• A PID has different sensitivities to different compounds. This is known in the art as a response factor or “RF”. Often times a PID will be calibrated on a specific gas, such as isobutylene for example, and the response factor of the PID to a particular compound will be referenced to its response to isobutylene. Response factors will vary slightly from one PID design to another. For example, a typical PID response factor for acetic acid is 11. This means that the PID’s response to 1 part per million (ppm) of isobutylene is 11 times that of its response to 1 ppm of acetic acid. Therefore, when such a PID is exposed to 1 ppm of acetic acid, it will read .09 ppm in units of isobutylene. In the art, this would be described as a reading of “.09 ppm as isobutylene”. A response factor influences the sensor’s ability to detect a compound at a given threshold. Detection will be most limited for compounds which have a combination of very low TLV or odor thresholds and very high response factors. In the case of acetic acid, which has an odor threshold of .016 ppm, it would not likely be detected by the PID in this example at its odor threshold, because this would be a reading of (.016 ppm / 11) .0014 ppm as isobutylene, which is beyond the resolution of most PIDs. When applied in conjunction with an exhaust fan monitoring application however assume for example thatPID401 is used to detect against a contaminant threshold of 0.4 ppm as isobutylene. In this case, when exposed to enough acetic acid to produce a reading of 0.4 ppm as isobutylene 4.4 ppm of acetic acid will be present at the sensor location of401 inplenum307. In order to ensure that the odor of acetic acid will not be present at a receptor point around the building, theexhaust fan system400 will have to provide 275:1 dilution. This is usually achievable even at lower fan exit velocities.
• The use ofsensor401 to monitor for exhaust contaminants has several drawbacks. First, it is often the case that theexhaust air314,315,316, and317 does not mix in a uniform manner withinplenum307. This often results in different contaminant concentrations being exhausted by each fan (309,310, and311). As a result, there often is no single good location within307 to apply asensor401 which would yield a contaminant measurement that’s sufficient to ensure fan exit velocities are properly regulated. For example,sensor401 may be placed on one side of the plenum near Riser 1 (302) as shown, however, this may not be sufficient to detect contaminants traveling through Riser 4 (305). This is especially the case ifonly fans309 and311 are active, in which case, contaminants inRiser 4 may not be detected at all, as theflow317 travels straight throughplenum307, directly intointake308 and throughfan311. In such a scenario, a potentially dangerous condition would result in which thefan system400 would continue to operate as a lower exit velocity in which contaminants may not sufficiently be dispersed from the building exhaust. This can result in entrainment of contaminants at unhealthy levels into building ventilation intakes or to other sensitive outdoor receptor locations.
• Another factor which can make the use ofsensor401 unreliable is that, attimes sensor401 may be exposed to very high concentrations for extended periods (many hours in some cases) and this has a tendency to foul the sensor. Note that even though exhaust air such as221 is often quite clean, at times it may in fact be quite rich with contaminants. This for example could occur during some period of time during portions of the week in which fume hood use is prevalent. Sensor fouling is especially an issue with PID’s when for example their lamp is exposed to high contaminant concentrations that can result in the buildup of a contaminant film that alters the lamp’s UV output intensity, often resulting in a sharp decrease in the sensor’s sensitivity. The end result is that, even after a few days of exposure,sensor 401′s ability to detect compounds with sufficient sensitivity will be compromised. PID’s and most other sensors are not designed for constant exposure to the high concentration of compounds which at times will be present from theClass 4 exhaust air that flows from fume hoods and other exhaust sources into total system exhaust221.
• FIG.4 also illustrates an alternative prior art method of detecting concentrations of contaminants in the exhaust air221, usingsensor402, which is integrated within a type of multipoint air sampling system known as a networked air sampling system.Sensor402 will at least comprise a PID, but may also include other sensors including a sensor to measure airborne particulate matter and a metal oxide semiconductor (MOS) sensor for measuring some VOC’s, such as methyl alcohol which the PID cannot sense. The prior art system shown incorporates asensor suite403, an air router (411) and foursampling locations414,417,420, and423. It should be clear that this example only uses four risers but that the multipoint air sampling system could be adapted to monitor more locations if necessary. The multipoint air sampling system shown inFIG.4 is described in U.S. Pat. No. 6,425,297 B1, which is incorporated herein. Using this system, air samples are conveyed to402 through acommon backbone tubing409, which is connected to theriser locations302,303,304, and305 throughvalves412,415,418, and421 housed withinAir Router411. Air samples are drawn toAir Router411 throughduct probes414,417,420, and423 viatubing413,416,419, and422 in a sequential manner in order to obtain discrete measurements of eachlocation302,303,304, and305 in a time-multiplexed manner. The sampling sequence is commanded via a centralized server (427) that may communicate with a plurality ofSensor Suites403 measuring different locations in the building.Sensor Suite403 communicates with one ormore Air Routers411 via acommunications network408 to instruct411 of whichvalve412,415,418, and421 should be open at a given time in order to facilitate an air sample. Air samples are drawn through theAir Router411 andSensor Suite403 via avacuum pump404. Control logic within Information Management Server427 compares the contaminant level sensed by402 against a pre-determined threshold in order to establish whether or not the fan system should operate at reduced exit velocities or not. When a PID sensor is used, it is common to set this threshold to between 0.2 and 1 ppm as isobutylene. This information is communicated either directly to the fan system or through the BAS vianetwork424. In manycases network connection424 will be a BACnet network connection, such as BACnet/IP. Those who are experienced with HVAC control systems will recognize that BACnet is a universal networking and communication protocol that was established by ASHRAE in order to enable different systems to communicate without the need for a proprietary network protocol. Also, not shown inFIG.4,Air Router411 has the ability to support analog signal connections to other systems. This includes the ability to provide a relay contact which may be monitored by another system in order to convey a binary or two-state condition. This and other signals (also, generally known in the art as I/O) can be provided through theAir Router411, which could directly connect to the controls which operatefans309,310, and311. Thus, logic that runs on Information Management Server427 will determine whether the fans need to operate at high or low exit velocities based upon whether contaminants have been detected in each of the risers or not, in order to create an “enable” signal which can either be communicated through theSensor Suite403 to I/O on theAir Router411 or the enable signal can be communicated via424 over BACnet to the BAS or fan controls.
• Information Management Server427 also has the ability to communicate contaminant levels sensed bysensor402 to aremote data center426 via aninternet connection425. In addition, basic diagnostic information on the operation ofsensor suite403 is provided through thisinternet connection425 to thedata center426 in order to be able to remotely monitor overall system health.
• Using a multipoint air sampling system in conjunction withsensor402 has the advantage of enabling the detection of contaminants inexhausts314,315,316, and317 before they comingle inplenum307. This enables the detection logic (which may be setup in Information Management Server427) to discriminate at a higher threshold or contaminant concentration than would be possible by monitoring only one point inplenum307. Detection that’s based upon measurements taken from theindividual risers302,303,304, and305 can provide an extra margin of safety and sensor noise rejection by taking advantage of the internal dilution provided by the total system exhaust.
• The internal dilution of theexhaust system400 is a factor that can vary considerably depending on how much total system exhaust CFM there is in comparison to a contaminant source’s CFM, such as that of a fume hood. In most facilities the dilution level that may be provided to a chemical spill within a fume hood could easily be a factor of 30 or more. This internal dilution component lessons that amount of dilution that the exhaust fan needs to provide. For most chemical inventories, 3000:1 dilution from the exhaust system (including internal dilution and dilution from the fan system) is usually more than sufficient. There are exceptions to this; however, these are usually identified when a dispersion and chemical inventory analysis is performed. With an internal dilution of 30:1, the exhaust fan needs to deliver as much as 100:1 dilution. Most fan systems can easily provide such dilution, even at exit velocities which are less than 3000 fpm.
• Similar to the issue withsensor401, even though the contaminant concentrations inrisers302,303,304, and305 will typically be low, in most facilities (especially one’s with fume hoods) it is common to encounter high concentrations of contaminants in the exhaust streams for periods of time throughout a given week. This has the tendency of fouling thesensor402. This is especially the case as the concentrations seen in each riser will be far higher than that seen in theplenum307. Airborne contaminants within the exhaust in eachriser302,303,304, and305 can easily reach 50 to 100 times the toxic limit value (TLV) or odor threshold of particular substances. This would often be the case when there is a chemical spill within a fume hood.
• A disadvantage that the networked air sampling architecture has when applied to400 is that the sharedbackbone409 tubing tends to adsorb certain compounds such as ones which are highly polar in nature. Those who are experienced with molecular chemistry will appreciate that molecules which are polar in nature have a separation of charge that will cause them to interact with other molecules with dipole-dipole interaction as well as hydrogen bonding. The polarity of a compound can significantly affect the performance of a sample draw system (such as a multipoint air sampling system) that essentially must convey the compound through a tube over some distance. Generally, the longer thetube409 is, the more intense the adsorption problem will be. Because of adsorption the response of the sensing system can be inhibited in such a way that it can take more time for a complete response to be seen atsensor402 as a highly polar compound is conveyed throughtubing409. This effect can be exacerbated as alternating clean samples are sequenced through thesame tubing409. This is normally not a problem when applying a networked air sampling system to monitor typical occupant environments (which is what these systems are primarily designed for) because the chemical compounds that are usually found in those environments will be very different than the broad array of compounds that may be used in fume hoods.Risers302,303,304, and305 will typically contain air from a combination of fume hoods, lab space, and non-lab space.
• One characteristic of the networkedair sampling system400 is that it is designed to capture and record a measurement of contaminant concentration for eachlocation302,303,304, and305 in a deterministic manner and without interruption. Therefore, whethersystem400 has indexed the fan outlet velocities to a higher value (for example 3000 fpm) because contaminants have been detected does not alter the sampling sequence of the system.
• One measure of a compound’s polarity is what’s known in the art as dipole moment. Dipole moment is measured in units of Debyes. Generally, a comparison can be made from one compound to another as to the degree to which they may adsorb totubing media409 by comparing dipole moment data.Tubing media409 includes but is not limited to Kynar^® and other fluoropolymers. For example, ammonia, which has a dipole moment of 1.47 Debyes, has a high tendency to adsorb to Kynar^®. Benzene, which has a dipole moment of 0 Debyes, has very little tendency to adsorb to Kynar^®. Other factors come into play which influence compound adsorption, such as what’s known in the art of molecular chemistry as van der Waal forces, but measurement of dipole moment can serve as a good indicator of compound adsorption tendencies.
• Another disadvantage of the networked air sampling architecture shown in400 is that it is relatively expensive to implement due to the number of components such asAir Router411,Sensor Suite403, and Information Management Server427 which have to be installed in addition totubing409,422,419,416,413, andnetwork connection408. The network air sampling architecture is more suitable for monitoring numerous occupant locations in a facility (usually several dozen locations) and can be expensive in terms of material and hardware costs along with installation costs to be justified for use with exhaust demand control applications.FIG.4 illustrates an application where only fourlocations302,303,304, and305 must be monitored. Often times the number of locations to be monitored may be even less, depending on the number of risers. In such cases where few locations need to be monitored in order to implement exhaust demand control, the application of a networked air sampling system would be far too complex and expensive to be practical.
• Historically, exhaust demand control sensing methods which eitheruse sensor402 within a multiplexed air sampling system ordiscrete sensor401, have been based upon a continuous monitoring approach. This means that themethod401 or402 continuously provides sensing, regardless of the contaminant concentration levels that are present in the lab exhaust (314,315,316,317). The continuous monitoring approach results in an exhaust demand control functionality that is often overly responsive when exhaust contaminant levels vary rapidly and by large concentrations. Rapid and pronounced variations in lab exhaust contaminant levels is something that’s quite common in many exhaust systems, especially as a result of chemical use in fume hoods. For example, many liquid organic solvents with high vapor pressures are often used in liberal quantities within fume hoods such as205 and208 by researchers. Because of these high vapor pressures, these solvents will vaporize readily and in high concentrations, due to spills or intentional releases involved with a process. Solvents may be used for example as a part of a liquid chromatography process, which may result in frequent releases of solvent vapors at high concentrations. Some examples of common high vapor pressure solvents include: toluene, hexane, dimethylformamide, tetrahydrofuran, isopropyl alcohol, ethanol, and many other solvents. Depending on the process, these vapors may be released over intervals as frequent as several minutes or less. Such solvent vapor releases withinfume hoods205 or208 for example can result in rapid and pronounced variations of solvent concentrations within system exhaust221 which would be detected by the exhaust demand control sensing method. This can result in oscillations or hunting in theexhaust fan systems300A or300B which can affect the flow control stability of exhaust221 and seriously affectfan203 and related equipment service life.
• Another type of multipoint air sampling system that might be applied to monitor contaminant levels in eachriser302,303,304, and305 is what’s known as a star or “hydra” configuration. With this type of system, sequenced air samples from multiple locations are brought to a common panel or suite which contains the sensor(s) and the valves which are used to capture air samples. A star configuration multipoint air sampling system can be similar to aspects of a networked air sampling system, such as that depicted inFIG.4, in that it incorporates many of the components of arouter411 andsensor suite403, typically within one enclosure. A star system does not incorporate abackbone tubing409. Examples of star type systems include but are not limited to the HGM-MZ by Bacharach Inc. and the MultiGard™ 5000 by Mine Safety Equipment Inc. These systems incorporate a variety of infrared sensor technologies, such as photoacoustic infrared spectroscopy and various non-dispersive infrared (NDIR) approaches for monitoring, in some cases total organic compound (TOC) content, but usually to target specific refrigerants. TOC sensing such as this generally provides sensing of fewer TVOC parameters than can be detected by a PID sensor. These systems are designed primarily for refrigerant leak detection which involves monitoring occupant breathing zones in buildings. They are not designed to monitorClass 4 or evenClass 3 environments, and their sensor technology can be more prone to fouling than a PID sensor due to their optical sensor technology. This is especially an issue with NDIR systems. Depending on the design, some photo-acoustic sensors (PAS) can be robust against fouling but, this technology is more appropriate for speciating one or a few compounds and not a broad simultaneous detection of the range of compounds found inClass 3 and 4 environments. Star-configuration multipoint sampling systems such as this are generally much lower in cost to implement than a networked air sampling system.
• The tubing media such as fortubing413,416,419, and422 can vary based on application and commercially available systems. With star configured multipoint sampling systems, such as most refrigerant monitoring systems, high density polyethylene tubing (HDPE) is commonly used. For systems which need to also monitor particulate matter in addition to volatile organic compounds and other parameters, an electrically conductive tubing may be utilized. U.S. Pat. No. 7,360,461 B2 describes a structured cable used with a networked air sampling system that incorporates power and communications wiring with an electrically conductive Kynar^® tubing that is doped with carbon nanotubes in order to achieve good electrical conductivity and inertness to chemical exposure.
• Although many exhaust fan systems such as300A and300B are designed to incorporate abypass air element301, it is possible to specify these systems to be configured without anybypass air option301. This can be accomplished if the physical size of the exhaust fan (309,310,311,323, or more generally203) can be chosen so that there will be adequate fan exit velocity and dilution characteristics when the total system exhaust CFM (221) is at a design minimum value. When this can be accomplished, better energy efficiency will result because excess fan energy will not be expended onbypass air301. Such an approach can be problematic, however, if at a future date it becomes desirable to reduce lab flows which contribute to exhaust CFM221, as this can result in insufficient exit velocity ofair312. This has become a serious issue for example with many legacy lab systems that were first commissioned years ago using higher lab air change rates, based on what used to be acceptable practice. If for example, the labs ofFIG.2 were originally commissioned at 12 ACH (what used to be acceptable practice), the potential for saving energy by reducing lab ACH values to 6 ACH would be significant but not possible if the reduction of lab exhaust CFM221 would result in unsafe fan exit velocities.
• Over recent years, there have been tremendous advancements in the capabilities of smart device available for Internet of Things (IoT) applications. These applications focus on providing low cost ways to share data between simple devices (including sensors and electronics) and other Internet connected devices and systems. At the heart of these advancements has been the development of very fast microcontroller products with data processing capabilities that rival that of personal computers, while also being small in size and easy to integrate into an electronic design. The data communications (herein IoT communications) capabilities supported by these smart devices (herein IoT modules) is well known to those familiar with IoT technology.
• Note that IoT communications may incorporate methods of connecting data to the Internet or Internet connected servers, using one or more stages of the communications may not incorporate an Internet protocol. As an example, Sigfox is a popular cellular network used to indirectly connect devices to the Internet. As is known to those skilled in the art of IoT, Sigfox employs a proprietary technology using the ISM radio band to provide a low power wide reaching wireless connection. In this example, individual devices are not connected directly to the Internet, but are connected through the Sigfox network to the Internet.
• One of the IoT communications known in the art is LoRa^® and is described in U.S. Pat. No. 7,791,415 which is incorporated herein. LoRa^®, which stands for Long Range, is a low power wireless communication technology that is well suited for transmitting data over great distances within almost any building. Historically, wireless communications within most buildings over more than a few hundred feet distance is problematic for most other wireless communications. LoRa^® is suitable for communications between devices located in a building, but it also provides an effective way to connect devices to the Internet.
• Other IoT communications include but are not limited to: WiFi/IEEE 802.11, Bluetooth, Bluetooth Low Energy (BLE), Sigfox, 6LowPan, IEEE 802.15.4, Ethernet, LPWAN, MQTT, Thread^®, and a number of cellular technologies (LTE CAT M1, 2G, 3G, 4G).
• Examples of IoT modules include but are not limited to: various modules by Particle IO (Photon, Electron, Xenon, Boron, Argon), ESP32 by Espressive Systems,Raspberry PI 3 Model B by Raspberry Pi Foundation, IMP005 by Electric Imp, Arduino MKR1000 by Arduino, various modules by Pycom Inc. (Lopy, Fipy, Sipy), and PIC-Web by Olimax Ltd.
• Other methods of connecting devices or systems to the Internet have existed for many years. One of the more relevant methods has included a physical Ethernet connection using what’s known in the art as a gateway, router, or server (herein server). In building automation applications, the server often consists of an industrial grade computer which may run one of any number of operating systems, including but not limited to any version of Microsoft^® Windows, Windows Server, Linux, and other operating systems. The server will often run software that’s known in the art as a “service”, and said service is responsible for collecting data from the devices at its location and communicating this data to the Internet. Often this Internet communications is accomplished using what’s known in the art as TCP/IP Sockets communications. Said sockets communications communicates over the Internet with another service that is also running on a server at another location. For example, information management server427 is an example of such a server that would operate the described service. An example of a product which operates in this manner is the IMS100, which is manufactured by Aircuity, Inc. Communications from a field device using a server such as this shall also be considered to be IoT communications.
• Examples of IoT modules include but are not limited to: various modules by Particle Industries Inc. (Photon, Electron, Xenon, Boron, Argon), ESP32 by Espressive Systems Pte. Limited,Raspberry PI 3 Model B by Raspberry Pi Foundation, IMP005 by Electric Imp^® Inc., Arduino MKR1000 by Arduino S.R.L., various modules by Pycom Limited (Lopy, Sipy, Wipy).
SUMMARY
• Embodiments of the present invention provide systems and methods which enable the reliable implementation of exhaust demand control for the energy efficient operation of a laboratory exhaust fan system. Aspects of the invention address elements to enable the reliable sensing of lab exhaust contaminants using a multipoint air sampling system. Embodiments of this invention apply to any type of multipoint sampling system, including but not limited to star configuration and networked air sampling systems. While exemplary embodiments are shown in conjunction with a PID sensor, the methods are suitable for any type of sensor that can be used with a multipoint air sampling system.
• In one aspect of the invention, one or more measures to ensure sensor accuracy and reliability includes methods of isolating the one or more sensors from contaminants when sensed contaminant levels exceed an action level setting. Another aspect of the invention ensures sensor accuracy and reliability using methods of sample dilution. Still another aspect of the invention ensures sensor accuracy and reliability by applying methods of flushing the air sampling tubing connected from each sensed location and the multipoint air sampling system, said flushing is applied when sensed contaminant levels exceed an action level setting.
• In other aspects of the invention, the exhaust demand control logic includes embodiments which ensure the stable operation of the exhaust fan system as it is commanded in and out of a state of setback. This includes embodiments which incorporate a fixed sequence delay, as well as embodiments which incorporate an adaptive variable delay within the control logic.
• Embodiments of the invention may also incorporate embodiments which both enhance the fail-safe and overall safety aspects of exhaust demand control, using a number of fan setback override features. An exemplary fan setback override feature is based on system error conditions. As an embodiment, a system error condition which will override fan setback, includes an override which is activated over an IoT connection when it is determined that the calibration of one or more sensors of the multipoint air sampling system has expired. Other embodiments include the use of an occupancy signal to override fan setback. In this embodiment and example application includes certain conditions where high risk lab chemistry may be in use. Still other embodiments include overriding fan setback when certain weather conditions exist in which it may not be beneficial to operate the exhaust fan system at reduced exit velocities.
• Aspects of this invention may be suited for use with exhaust fan systems that incorporate a bypass air element. In one embodiment, the bypass air element is modulated so that the total CFM delivered by the exhaust fan system is reduced when commanded into a state of setback. For other cases in which a bypass element does not exist within the exhaust fan system, embodiments of this invention provide an exhaust demand control function which incorporates clean exhaust minimum ACH logic.
BRIEF DESCRIPTION OF THE DRAWINGS
• The foregoing features of this invention, as well as the invention itself, may be more fully understood from the following description of the drawings in which:
• FIG.1 is a simplified schematic illustration of a prior art recirculating air distribution system;
• FIG.2 is a simplified schematic illustration of a prior art multiple zone lab ventilation system;
• FIG.3A illustrates a conventional prior art high plume fan system;
• FIG.3B illustrates a conventional prior art exhaust fan system using a fan system other than a high plume fan system;
• FIG.4 illustrates two prior art approaches used to detect contaminant levels in a lab exhaust stream;
• FIG.5 is a flow diagram which illustrates elements of the exhaust demand control logic and other functions in accordance with embodiments of the invention;
• FIG.6 is an illustration of aspects of an exhaust demand control system in accordance with embodiments of the invention;
• FIG.7 illustrates embodiments of the exhaust fan setback override in accordance with embodiments of the invention;
• FIG.8 illustrates embodiments which contribute to the sensor performance in accordance with embodiments of the invention;
• FIG.9 illustrates additional embodiments which contribute to the sensor performance aspects in accordance with embodiments of the invention;
• FIG.10 is a detailed illustration of the valve logic of a multipoint air sampling system in accordance with embodiments of the invention;
• FIG.11 is an illustration of embodiments involving control logic used to lower exhaust fan flows, as elements of the exhaust demand control logic in accordance with embodiments of the invention; and
• FIG.12 is a schematic representation of an exemplary computer that can perform at least a portion of the processing described herein.
DETAILED DESCRIPTION
• Embodiments of the invention provide apparatuses and methods for exhaust demand control. Embodiments of the invention are useful for monitoring contaminants within at least a portion of the exhaust air conveyed by an exhaust fan system in order to control one or more aspects of the exhaust fan’s energy use. Reliability and ease of implementation of contaminant sensing are addressed in illustrative embodiments of the invention. In some embodiments, a system can factor in variable outdoor environmental conditions in order to influence the reliability of the exhaust demand control. A yet further aspect factors variable indoor conditions to minimize risks associated with the performance of the exhaust demand control. Example embodiments are applicable to high plume fans or other lab exhaust fan systems which may incorporate a bypass at one or more locations within the exhaust ductwork or plenum in order to influence the fan’s dilution or exit velocity characteristics. In addition, some exhaust fans do not incorporate a bypass but can vary the amount of dilution they provide along with dispersive characteristics by varying fan speed. For example, entrained air can be increased at the fan’s nozzle in order to provide added dilution to the total system exhaust.
• Sensors used to detect indoor air contaminants may be exposed on a continuous basis to the environment or environments being sensed by the device. Data from these sensors may be recorded on a fixed time interval using well established sampling techniques or the data may be generated on an irregular basis, using a change of value (COV) technique. COV based monitoring records or communicates data when the measured property has changed by some predetermined amount from its last recorded value. In the case of a discrete sensor such as401, it is continuously exposed to the air withinplenum307, regardless of the intensity of contaminants there within. By comparison,sensor402′s exposure to exhaust contaminants may on average be similar to that seen bysensor401 however, it will tend to see higher peak contaminant concentrations than401 because individual air samples fromrisers302,303,304, and305 are conveyed tosensor402 in a sequential manner. These peak contaminant levels may foulsensor402 more readily than the rate at whichsensor401 is fouled.
• With the exhaust demand control application, the exhaust air should be examined for contaminants with some minimum frequency. For example, when there is a chemical spill within a fume hood that is served by an exhaust fan system the exhaust demand control strategy should be able to detect the presence of contaminants at concentrations which exceed a pre-determined threshold (herein the action threshold) in order to command the exhaust fan to deliver a maximum exit velocity within a few minutes (e.g. 2 -3 minutes) of the spill. This means that, for those times where exhaust contaminants are less than the target threshold, samples fromsensor401 need to be taken every few minutes. Forsensor402,location302,303,304, and305 should be sampled within the same few minute period. When exhaust contaminants exceed the action threshold value however it is not necessary to continue to sample data fromsensors401 or402 at the same rate, and a much slower sampling rate is possible while ensuring safe operation of the fan.
• As a measure to ensure sensor accuracy and reliability, an embodiment of this invention takes advantage of the fact that when exhaust contaminant levels exceed the action threshold, as long as the exhaust fan has been enabled to operate at a higher exit velocity, data from the sensor used for exhaust demand control can be acquired at a reduced rate and therefore, the average exposure duration of the sensor may be reduced in order to protect the sensor from fouling.
• Example embodiments incorporate a multipoint air sampling system designed to monitor exhaust air which may include one to any number of sampled locations; herein referred to as monitoring points. In atypical application 1 to 6 monitoring points may be sufficient, however, embodiments of the invention may support any number of monitoring points. Generally, a monitoring point is required per exhaust duct riser that connects to theplenum307 of theexhaust fan system203 to which exhaust demand control is being applied. However, embodiments are not limited to monitoring only a single location per riser. In some cases, it will be beneficial to monitor multiple locations along the length of a riser, or even multiple locations that run horizontally on a given floor in the building. In some applications, such as when there are only limited number of fume hoods in a building, it may be advantageous to monitor locations in the exhaust duct where effluent from clusters of fume hoods is concentrated, rather than assigning monitoring points to known clean exhaust sources. Although the figures included in this disclosure illustrate scenarios where a single monitoring point is applied per riser, example embodiments are not limited to a single monitoring point per riser, as it applies to the implementation of any number of monitoring points per riser and further does not require that all risers be monitored. In some embodiments, no monitoring points will be assigned to a riser but instead may be assigned to various horizontally connecting duct locations in the building. More generally, embodiments of the invention may apply to monitoring any exhaust duct location that may include riser locations, one or more locations within the plenum such as307, horizontal ductwork locations per floor, individual exhaust locations at a fume hood, as well as individual duct locations specific to canopy hoods, snorkel exhausts and general exhausts.
• FIG.5 is a flow diagram which provides illustrative embodiments of the exhaust demand control logic which includes air sampling logic and other logic.FIG.6 is an illustration of further details of an exhaust demand control system that includes a multiplexedair sampling system611 to which500 is applied.Logic500 is executed using one or more CPU’s that performcontrol logic605.Logic500 is therefore a subset of theoverall logic605 needed to operate thesystem611. In one embodiment, the one or more CPU’s which performlogic500 are contained within the panel or enclosure that houses theair sampling system611. In another embodiment, any portion of thelogic500 is executed by a CPU that is physically separated fromair sampling system611. As an embodiment,logic500 is performed by a CPU that communicates to611 through a data communications network which sends commands to611 as it executeslogic500. In one embodiment, the said data communications network is an internet connection to a cloud-based CPU, such as that contained within a server device. In another embodiment, the said data communications network is a local area network or LAN, said LAN includes any number of physical networks used to network devices in a building. In another embodiment, the said data communications is a point to point wireless connection betweensystem611 and the device with a CPU that is physically separated from thesystem611. In one embodiment, at least a portion oflogic500 is executed using a CPU that is part of the Building Automation System. In another embodiment, at least a portion oflogic500 is executed by an Internet connected CPU.
• Control logic605 controls the flow control components within610, as described further below.Control logic605 also acquires readings fromsensor602 and is responsible for any external communications such as624 (to the fan controls or BAS) andIoT Communications625 to a remote data center. In one embodiment,air sampling system611 is contained within one panel or enclosure and the one or more CPU’s which performcontrol logic500 and605 is executed using a microcontroller. Other embodiments can also include distributed applications of611, such as with a networked air sampling system topology where elements of611, such asvalves621,618,615 and612 along withvalve control410 may be located remotely fromcontrol logic605. In one embodiment,control logic605 may in whole or in part be contained remotely from valves (621,618,615,612),flow control610, andsensor602.
• As an exemplary embodiment, the CPU incorporated within605 is an ARM Cortex M3 micro-controller which incorporates a Broadcom Wi-Fi chip. Examples of this include the Particle Photon, by Particle Industries Incorporated, which is an Internet of Things (IoT) module. In this embodiment,control logic605 uses a Wi-Fi connection asIoT communications625 in order to accessremote data center626.
• As an embodiment, the provision ofIoT communications625 which connectscontrol logic605 andremote data center626, enablessystem600 to be remotely and proactively monitored by a support team or remote monitoring software that is part ofremote data center626, or both a support team and remote monitoring software. This enables issues withsystem600 to be identified and communicated to field personal who can address problems with600 as they occur. For example, said proactive monitoring can be valuable for identifying conditions that would result inerror condition705, such as but not limited to a failedvacuum pump627, which would causesetback signal517 to be set to False, thus causingfans309,310,311 to not be setback which results in higher energy consumption.
• One aspect oflogic500 is the ability forerror condition705 to provide enough functionality to ensure that the likely sources of electronics, power, or other failures insystem611 will not create a safety issue with theexhaust fan203. For example, a malfunction withsensor602 could result in an erroneous and potentially dangerous condition where fan setback signal is set to True, resulting inlow fan203 exit velocities, even when the exhaust221 contains high contaminant levels. Such a condition can result in environmental health issues for the building occupants. If forexample sensor602 is a PID sensor, it requires periodic maintenance and calibration (the calibration period), usually every 6 months for example. When thesensor602 in this example has operated for more than 6 months, it may be said that the calibration has expired. Such maintenance requires that a trained technician visit the facility in whichsystem611 is installed. If, however, that maintenance is not performed at theright interval sensor602 may not perform correctly, leading to anerroneous fan203 setback condition. In one embodiment, a sensor maintenance override may be activated throughremote data center626. In this embodiment,remote data center626 enables anerror condition705 to be set remotely so thatfan setback override701 is set to True and thefan setback signal517 is set to False, when the sensor calibration of602 has expired. This would be accomplished via IoT communications35, either using logic that is manually set withinremote data center626, or by using a programmed schedule within626. Using this approach of managing a sensor maintenance override using aremote data center626, as one embodiment, also enables an organized communication of this event to facility personnel and other individuals responsible for theexhaust fan system600, using email, texting, or social media such as but not limited to Facebook and Twitter.
• In embodiments,sensor602 can include one or more contaminant sensors including but not limited to: a photoionization detector, a sensing instrument based on photoacoustic infrared spectroscopy, a TOC sensor, an acid gas sensor to detect any of various acids, an airborne particle counter, an ammonia sensor, an arsine sensor, a chlorine sensor, a chlorine dioxide sensor, a combustible gas sensor, a diborane sensor, an ethylene oxide sensor, a fluorine sensor, a metal oxide semiconductor (MOS) sensor, a hydrazine sensor, a hydrogen chloride sensor, a nitric acid sensor, a hydrogen cyanide sensor, a hydrogen selenide sensor, a hydrogen sulfide sensor, a mercaptan sensor, a nitric oxide sensor, a nitrogen dioxide sensor, a phosgene sensor, a phosphene sensor, a silane sensor, a sulfur dioxide sensor, and a tetrahydrothiophene sensor. In one embodiment,sensor602 comprises a flame ionization detector (FID). FIDs operate on a similar principle as PIDs except, instead of utilizing a UV lamp to ionize compounds, an FID utilizes a flame to provide ionization via combustion. FID typically use hydrogen as the fuel source for the flame. An FID has the advantage of being able to ionize and therefore detect more compounds that are generally detectable by a PID. As one embodiment,sensor602 is a PID sensor. As an exemplary embodiment,sensor602 is a PID with a 10.6 eV lamp. Based upon the aforementioned sensor types which may be used forsensor602, it should be clear that embodiments can apply to either monitoring a specific compound or, a specific species, as well as to the application of broad sensor technology (such as a PID for example) which is not specific and does not speciate.
• Thelogic500 may apply to any suitable multipoint system topologies including but not limited to star configurations and networked air sampling systems. In one embodiment, the sequence is representative of the hardware and software which form the workings of an example multipoint sampling system which supports the exhaust demand control application when operating in conjunction with a high-plume exhaust fan.
• At the start of thesequence501, thesystem500 undergoes initialization where settings such as the number ofmonitoring points502 are loaded into the memory associated within a CPU contained within500, such as withincontrol logic605. As an embodiment, each time thesystem500 is reset, the fan (such asfans309,310, and311) will be commanded by500 to its maximum exit velocity vialogic504 for safety purposes, until500 establishes that the contaminant levels for monitoring points below theaction level515. The signal used to command said exit velocity state shall herein be referred to as a “fan setback signal”. As one embodiment, when thefan setback signal517 is “True” it conveys to the fan controls that the fan system should setback to a predetermined exit velocity. In this embodiment, when thefan setback signal517 is “False” it conveys to the fan controls that the fan system should operate at its higher exit velocity.
• As one embodiment, the action level may be a setting that resides within the field installed system, such as a value programmed into611 or as an alternate embodiment, it may be an electronic setting, such as a potentiometer or some other hardware setting within611. As yet another embodiment,action level515 is a value communicated to611 by an external device viainterface624 that includes but is not limited to: a building automation system (BAS), a networked air sampling system, a wireless or wired connection from a handheld device such as a mobile device. As another embodiment, the action level may be commanded or altered via thedata center626, or by way of what’s known in the art as a RESTful interface or API.Interface624 includes but is not limited to a BACnet network connection, an 802.11 or Wi-Fi interface, a Bluetooth^® or 802.15.1 connection, a Modbus network connection an RS485 communication network, a ZigBee wireless network, analog signals including but not limited to 0-10VDC or 4-20ma current loop. Note that theaction level515 will vary based on the application and the type of sensor used in602. If for example602 is a PID with a 10.6 eV lamp, then the action level may be set but is not limited to a setting between 0.4 and 1 ppm as isobutylene.
• It should be apparent to those skilled in the art of integrating HVAC controls equipment that any form of wireless or wired analog or digital communications can be used to supportinterface624. As one embodiment,interface624 also may support one or more relay contacts used to command the controls to theexhaust fans309,310, and311 to a higher lower exit velocity state based on the state ofsetback signal517. This embodiment has the advantages of providing electrical isolation betweensystem611 and the fan controls orBAS407, while also providing a signal that can be configured to be failsafe to a power outage at611. This is accomplished by configuring the relay of this embodiment so that a fan setback command fromlogic605 tocontrols407 is provided when the relay is in its energized state. For example, if a relay contact closure signifies a fan setback command the relay would be configured so that it needs to be energized for it to be in this state. Therefore, if power is lost, the relay will automatically be deenergized which will cause the contacts to open, thus signifying tocontrols407 that thefan203 should not be setback.
• Logic element504 also sets a counter variable “N” so that the air sampling sequence will start at the first monitoring point. Which monitoring point (614,617,620,623)system500 draws an air sample from first, is arbitrary and it should be clear that any order with which the sampling process acquires air samples from monitoring points is considered to be within the scope of this invention. Counter variable N withinlogic element504 is used to keep track of how many of the monitoring points (614,617,620,623) have been sampled during each cycle of the sampling sequence. One complete cycle occurs when all of the monitoring points (614,617,620,623) have been sampled. It should be clear that the number of monitoring points is not limited to 4, such as in this example, but that it can include one to any number of monitoring points.
• Inlogic element505, a sample is acquired from the monitoring point designated by counter variable N. For example, if N = 1 then an air sample from the first monitoring point in the sequence will be conveyed tosensor602, as described bylogic element506, usingflow control610.Flow control610 may incorporatevalve control410 andflow control428 and may incorporate capabilities to support sensorprotective mode509 which is discussed further below. As each sample is sensed bysensor602, as described bylogic element506, the measured contaminant value is stored in memory withincontrol logic605. The contaminant level measured inlogic element506 is then compared against the action level inlogic element507. If the contaminant level measured from the monitoring point in506 is greater than theaction level515, then the exhaust fan must be set to its higher exit velocity. This is accomplished inlogic element508 by setting thefan setback signal517 to “False”. This information is communicated via624 to the fan controls or the BAS which may be controlling the fan.
• Once a condition has been detected where contaminant concentrations exceed the action level causing508 to setfan setback signal517 to False, as an embodiment, multiplexedair sampling system611 will be placed into sensorprotective mode509. Sensor protective mode, a measure to ensure sensor accuracy and reliability, includes a number of embodiments which are intended to protect the one ormore sensors602 from fouling, drifting in calibration, or other forms of sensor malfunctions, and other influences which can cause thesensor602 to not read correctly as a result of exposure to high concentrations of contaminants in the exhaust streams314,315,316, and317 for extended periods of time. Embodiments are not limited to the number of monitoring points it can connect to and therefore, are not limited to monitoring just four exhaust streams such as314,315,316, and317. More generally, embodiments are applicable for monitoring from one to any number of exhaust streams. Whilesystem611 is in sensorprotective mode509, thefan setback signal517 will be set to False, resulting in the fan operating at its higher exit velocity setting for safety.
• As one embodiment of sensor protective mode, when this mode is enabled, multipointair sampling system611 will discontinue its air sampling sequence for a period of time designated bysequence delay503. By discontinuing the sampling process in611,sensor602 is isolated from the contaminants in the exhaust streams being monitored, which prevents the sensor from being overexposed on a continuous basis and thus ensures sensor accuracy and reliability will be maintained. As one embodiment,sequence delay503 is a fixed value. Typical fixed values ofsequence delay503 include but are not limited to values that range from 10 minutes to 20 minutes. In an embodiment,sequence delay503 may be a configuration parameter ofmultipoint sampling system611 that is set within thecontrol logic605 memory or that is based upon a hardware setting in611 that includes but is not limited to a potentiometer, jumper, or dip switch setting. As an alternate embodiment,sequence delay503 is a value that is communicated to611 viacommunication624 from the fan control system orBAS407.
• As an alternate embodiment,sequence delay503 is variable or adaptive depending on the frequency with which the sensorprotective mode509 is enabled. In this embodiment added protection can be provided tosensor602 by further reducing the frequency with which it is exposed to exhaust contaminants if it is found that the contaminant levels are above theaction level515 for an extended period of time. For example, when thesystem611 first goes into sensorprotective mode509, the sequence delay may initially be 10 minutes in duration. If after 10minutes system611 still measures contaminant levels above the action level the sequence delay may be increased further to 15 minutes for example. If after this 15-minute period system611 still measures contaminant levels above the action level the sequence delay could be increased further to 20 minutes, and so on. In this embodiment of an adaptive sequence delay an upper limit to the adaptive sequence delay may be defined in order to limit unnecessary exhaust fan energy use that could result for example at the end of a day where there was a lot of lab activity for an extended period of the day. For example, fume hood use in some facilities may be continuous for 4 to 6 hours of a working day, thereby potentially making the exhaust stream (314,315,316,317) contaminated above the action level for that period. In that case, anadaptive sequence delay503 that is not properly limited may result in a delay that is several hours long that would cause the exhaust fan to continue to operate at a high exit velocity, wasting energy, for several hours at the end of the 4 to 6 hour working period where the fume hoods are active. It may therefore be advantageous to limit the sequence delay to, for example, less than one hour.
• One advantage relates to sensor protective mode and to the stability of the control of the exhaust fan system when changing the exit velocity. As has been described, the exit velocity of exhaust fan systems (which includes high plume fan systems) is controlled using adjustments to thebypass air301 which may be accomplished using a static pressure control loop that involves controlling thebypass air301 in order to maintain a predetermined static pressure setpoint withinplenum307. This control may include proportional-integral-derivative control which results in a control loop that is robust for steady state operation but that has what’s known in the art of control systems as a “natural response” where the fan speed may temporarily oscillate in a dampened sinusoidal manner when sudden changes to fan speed are created. These oscillations may last for several minutes. In many exhaust fan configurations, the way in which an increase or decrease in exhaust fan exit velocity is achieved is by way of a change in fan speed, from which some level of fan system oscillation may be expected. For example, a reduction in exit velocity would start with a reduction in exhaust fan speed setpoint to each fan’s (309,310,311) variable speed drive (VFD). Typically, motor/fan speed is measured in Hertz (Hz), where zero Hz would infer that the fan is shut off and 60 Hz would be maximum speed. At maximum speed, the exit velocity and airflow delivered by each fan will be determined by the physical dimensions of the fan and the static pressure setpoint withinplenum307. A typical static pressure setpoint may be -4 inches H2O but that setting could vary considerably depending on the application. As the speed command to each VFD controlling the fans (309,310,311) is reduced in order to decrease exit velocities some amount of fan speed oscillation will result due to the natural response of the system. The same will occur each time each fan speed is increased. When exhaust demand control has been implemented, a common problem that is encountered is that speed control of the exhaust fan system can become unstable. This is what’s known in the art as “hunting” or “fan hunting”, which signifies that the fan system’s speed control does not reach a fixed steady state speed. Fan hunting can become a serious problem in that it can result in the premature failure of some fan components, such as the bearings in the fan assembly. The reason why fan hunting may take place with prior art exhaust demand control strategies is that it is often the case that contaminant levels in exhaust flow streams (314,315,316,317) fluctuate considerably above and below the action level over short periods of time (often 1 to 2-minute intervals). As a result, prior art exhaust demand control strategies can result in frequent changes to fan speed setpoint, thus resulting in system instability or hunting.
• In embodiments of the invention,logic500 can providesequence delay503, which not only protectssensor602 via sensorprotective mode509, but it also protects the fans (309,310,311) from excess wear and tear that would result from hunting. Such hunting is avoided because thesequence delay503 will often be set to 10 minutes or more, which is usually more than enough time for most fan speed changes to reach a fixed steady-state value and therefore the exhaust fan system will not hunt because thefan setback signal517 will not change as rapid changes to contaminant levels inexhaust streams314,315,316,317 occur.
• In an embodiment of this invention, settings withinlogic500 which include but are not limited to sequencedelay503 andaction level515 are established using potentiometers within the electronics which operate at least a portion oflogic500. In another embodiment, settings withinlogic500 which include but are not limited to sequencedelay503 andaction level515, are established as values in the memory associated with a CPU that performs at least a part oflogic500. In a preferred embodiment, the settings which may includedelay503 andaction level515 as well as other settings associated withlogic500 are configured using a local web page that is served by either a first CPU that at least performs a part oflogic500 or by a second CPU that is physically located within the same enclosure as said first CPU and that is in communication with said first CPU. As has been described, there are a wide variety of IoT modules available on the market and many of these modules have processing capabilities that is suitable for rendering a web page and most support some form of local communications, including but not limited to Blue Tooth and WiFi communications. In an exemplary embodiment, the settings which may includedelay503 andaction level515 as well as other settings associated withlogic500 are configured via a local web page that is served by an IoT module that is housed within the same enclosure as thesystem611.
• In embodiments, eachtime logic500 activates sensorprotective mode509 it resets the monitoring point counter “N” to 1 vialogic element511. Once thesystem611 has been in the sensorprotective mode state509 for the duration ofsequence delay503,logic element511 causessystem611 to reset so that the next monitoring point that it acquires a sample from vialogic505 is monitoringpoint 1 at the beginning of the sequence. Notice that as thesequence delay503 expires thefan setback signal517 will still be in the False state (as per logic508). This will be the case untillogic500 can successfully sequence through each monitoring point and confirm that the contaminant levels in each are below the action level. Following sensorprotective mode509,logic500 will loop516 andlogic element505 will acquiremonitoring point 1 and then that sample will be sensed and recorded vialogic element506. If the contaminant concentration in that first monitoring point is verified to be below the action level515 (via logic element507),logic element512 then verifies if the current monitoring point is then last monitoring point in the system. Given that in this example there are four monitoring points (via setting502) and that the current monitoring point is 1, thelogic500 will then proceed tologic element514 which then increments counter N by one, following whichlogic500 again loops via516 and the process continues. If contaminant levels in each of the four air streams (314,315,316,317) are found to be below theaction level515 vialogic element507, the logic will loop516 back throughpath505,506,507, and thenlogic element512 where, on the 4^th or final monitoring point, as verified bylogic element512, the logic path will be directed tologic element513 which then setsfan setback signal517 to True. Withfan setback signal517 set to True, exit velocities offans309,310,311 will then be setback viacommunication624 to the fan controls orBAS407. Followinglogic element513, the counter N is reset to 1 vialogic element511, and the sequence starts anew as it then loops again through516 and acquires a sample from the first monitoring point via505. Note that the target exit velocity at which the one or more exhaust fans (309,310,311) operate at when thefan setback signal517 is True or False may vary considerably from one application to the next. However, typical design exit velocities are 3000 feet per minute when the fan setback signal is set to True.
• FIG.7 depicts embodiments of this invention where a fansetback override function701 is implemented in order to further restrict the times where the exhaust fans (309,310,311) may be commanded to a reduced exit velocity. As shown inFIG.7, the setback override function provides an override input tofan setback signal517 and sensorprotective mode509, based upon a number of possible input conditions that may be monitored (703,704,705).
• When thesetback override function701 is set to “True” (the override condition) thenfan setback signal517 is set to False, preventing the exhaust fan from being setback. Simultaneously, sensorprotective mode509 will be enabled which will interrupt the air sampling sequence of611 and isolate and thus protectsensor602. In application, there are a number of conditions where it’s desirable to run the exhaust fan at a higher target exit velocity (such as 3000 feet per minute for example) even if the contaminant levels detected by602 inexhaust streams314,315,316,317 are relatively low and well belowaction level515. These conditions include but are not limited to certain occupancy conditions, certain weather conditions such as rain, and error conditions within611.
• As an embodiment, the fansetback override function701 may be a logic element within611 or it may be a logic element that is external to611. For example, function701 may exist within theBAS407 or some other external controller which communicates through424.
• As an embodiment,occupancy signal703 is a parameter that can be used for the determination of the fan exit velocity setting. This setting can be useful when it is desired for example to add an extra level of safety to the exhaust demand control application by not allowing the exhaust fan system to setback when certain portions of the building where611 is applied become occupied. For example, in one embodiment, occupancy sensing from certain laboratory locations where chemical and fume hood use may be possible when said laboratory locations are occupied could be used to create asignal703 which, when703 signifies an occupied condition,fan setback override701 will be set to True. As an alternate embodiment,occupancy signal703 may be generated from an occupancy schedule that is programmed into the BAS or other external system from611 or that is programmed intosystem611.
• Depending on the design offan309,310,311, it may not be desirable to setback to a lower fan exit velocity when it is raining outside. ANSI Z9.5 recommends a fan exit velocity of 2000 feet per minute or more may be required to prevent moisture from getting into the fan system, which can cause equipment malfunctions or even water migration into locations within the building.
• One embodiment uses fansetback override function701 and a weather/rain signal704 to prevent fan setback when it is raining outside. In one embodiment, signal704 may be derived from a rain sensor that is mounted in proximity to the building served by611.
• In this embodiment, rain sensor signal may be connected directly to611 or it may be read via theBAS407 or other remote device communication throughconnection624. As an alternate embodiment, signal704 may be obtained from local weather data that is communicated through internet or internet of things (IoT)connection625. In this embodiment, said weather data may be obtained through what’s known in the art as a RESTful interface to an application programming interface (API) provided by an internet weather site. Examples of such sites which offer API’s for collecting weather data include but are not limited to: weather.com, wunderground.com, theweathercomany.com and aerisweather.com.
• Another embodiment of this invention involvesinput705 to the fansetback override function701, which is based on error conditions insystem611.Error condition705 enablessystem611 to operate with an excellent level of fault tolerance by ensuring that if any number of error conditions associated withsystem611 arise, the exhaust fan system will not be allowed to setback to a lower exit velocity. Such error conditions include but are not limited to: a failure withvacuum pump627, a malfunction with any of thesensors602, detected blockages or malfunctions associated with any of the valves (612,615,618,621), malfunction withinflow control610, or other malfunctions that are detected withinsystem611.
• Each time through logic loop516, a sample from one of614,617,620, or623 is conveyed throughtubing613,616,619, or622, respectively. As embodiments, a number of tubing materials are suitable for this purpose, including but not limited to: high density polyethylene (HDPE), Kynar^®, and a number of fluoropolymers including Polytetrafluoroethylene (PTFE) and Polyvinylidene fluoride (PVDF). As an alternate embodiment,tubing613,616,619,622 are made of stainless steel, such as308 or316 stainless tubing. As a preferred embodiment, the tubing is made from Kynar^®, has an inner diameter of ⅛ of an inch and an outer diameter of ¼ inch.
• Sensorprotective mode509 includes embodiments in addition to protectingsensor602 by isolating it from exhaust contaminants.FIG.8 illustrates added embodiments withinflow control610, which are measures to ensuresensor602 accuracy and reliability.Function block802 incorporates the airflow regulation necessary to draw air samples in a consistent manner. This includes but is not limited to any type of airflow regulation device such as for example a mass flow controller, or what’s known in the art as a critical flow orifice, or a critical flow venturi. Those experienced in the art of air or gas flow control will appreciate that a wide range of approaches exist in the art that are applicable for use in multiplexedair sampling system611.
• Airflow control element610 inFIG.8 also incorporates agas flow device803, which is intended to provide a number of functions that relate toprotective mode509 operation as well as other operations which are beneficial to the performance ofsensor602, thus ensuringsensor602 accuracy and reliability.Flow device803 includes but is not limited to embodiments based on a solenoid valve, a flow orifice, or a mass flow controller. As an embodiment of this invention,flow device803 protectssensor602 by providing dilution sampling, which reduces concentrations of contaminants in a controlled manner. In this embodiment,flow device803 provides a flow of clean air which mixes with the air sample from a monitoring point which flows through802. Said clean air can include any contaminant free source. For example, it can include relatively clean ambient air, that would be considered clean in comparison toexhaust streams314,315,316,317. As an alternate embodiment the clean air source may be ambient air that is drawn through agas cleaning device804, which may incorporate any form of filtration including but not limited to activated carbon, molecular sieve material, and particulate filtration media.Cleaning device804 will be selected based upon the sensor elements contained within602 but, as a preferred embodiment, will usually include media that can remove volatile organic compounds (VOC’s) from the ambient air.
• In the dilution sampling embodiment,flow device803 provides a controlled source of clean air that is void of the target gas sensed by602 in order to reduce the exposure ofsensor602 to that gas. This measure further ensuressensor602 accuracy and reliability. For example,clean air source803 may be adjusted bycontrol610 so that the clean air flow from803 is delivered at a fixed percentage of the total airflow rate delivered tosensor602. As a further example, if the airflow rate delivered tosensor602 is 2 liters per minute andflow device803 is adjusted to deliver 1 liter per minute, then the air sample delivered throughfunction block802 will be diluted by 50%. This reduces the maximum exposure seen bysensor602. Continuing on this example, if theaction level515 of contaminants is .4 ppm (as isobutylene), then the dilution of 50% ensures that the maximum exposure ofsensor602 will not exceed .2 ppm as isobutylene. In another embodiment, dilution sampling via802 reduces the exposure ofsensor602 to contaminants that it doesn’t sense. For example, inmany applications sensors602 may be a single PID sensor which mostly senses VOC’s and some limited number of inorganic compounds. An inorganic compound that it does not sense is nitric acid. Nitric acid fumes will not normally reach concentration levels in anexhaust stream314,315,316,317 that require high levels of dilution from theexhaust fans309,310,311, however, some low-level exposure of nitric acid over time can contribute to the fouling ofsensor602. By incorporating adilution sampling component802, it can dramatically reduce the exposure of that non-sensed parameter. In this embodiment,flow device803 provides a controlled source of clean air that is void of any gasses likely to be contained within samples taken from a monitoring point. This reduces the exposure ofsensor602 to gases that it both senses or does not sense. As an embodiment, instead of conveying clean air throughfilter804,flow device803 conveys a clean gas from a gas cylinder, which may include but is not limited to pure nitrogen gas, or a mix of nitrogen with oxygen (also known in the art as “zero air”).
• In another embodiment, which is a further measure to ensure sensor accuracy and reliability,flow device803 is enabled when sensorprotective mode509 is activated in order to provide a flushing function. In this embodiment when protective mode is activated,sensor602 only receives clean airflow from device803 (which in this embodiment may be a solenoid valve or some other airflow switching device) and no airflow is received fromflow device802 during this state. This provides a flushing action that desorbs contaminants from thesensor602 and its enclosure and tubing. In this embodiment,vacuum pump627 continues to operate even though air samples will not be conveyed from the monitoring points. In this mode,vacuum pump627 provides the suction to convey the airflow through803 and602. Over the course of operation of multiplexedair sampling system611, the adsorption of compounds or contaminants from exhaust air streams314,315,316,317 to the surfaces thatsensor602 is exposed to (for example:sensor602 enclosure, tubing, and other surfaces in the flow path) can result in low level desorption that alters the accuracy of thesensor602 readings. By flushing this flow path, it will minimize the buildup of adsorbed contaminants which would augment the accuracy of thesensor602 reading.
• In another embodiment to ensure sensor accuracy and reliability,valve801 is included withinsystem611. In this embodiment,valve801 is a three-way valve, such as a three-way solenoid valve. When thesystem801 is sequencing air samples from the monitoring points, such as flow streams314,315,316,317, three-way valve801 will provide a flow path between thevalves612,615,618, and621 through which each monitoring point sample is conveyed and flowregulation device802. As an embodiment, when sensorprotective mode509 is enabled, three-way valve801 interrupts this flow path and simultaneously provides a flow path between the common side ofvalves612,615,618,621 and ambient air. At that moment, in this embodiment,valves612,615,618, and621 are all commanded to their open position. Eachexhaust monitoring point314,315,316,317 is negatively pressurized, owing to the inherent function of theexhaust fans309,310,311. As was mentioned, typically theplenum307 to whichrisers 1,2,3, and 4 connect, is controlled to a fairly high static pressure, such as -4 inches H2O. As a result, withvalves612,615,618, and621 open andvalve801 open to atmosphere (ambient air) this embodiment enables relatively clean ambient air to flow throughvalve801 throughvalves612,615,618,621, and throughtubing613,616,619,622 where it exits into the negatively pressurized flow streams314,315,316, and317. This provides a flushing function totubing613,616,619, and622 that is advantageous as this action removes adsorbed compounds which setup in the tubing that can over time affect the accuracy of the contaminant readings performed bysensor602.
• FIG.9 illustrates an exemplarymultipoint sampling system900 used in embodiments of this invention.900 includes an exemplary embodiment ofairflow element610 which incorporates two airflow paths and measures to ensuresensor 602′s accuracy and reliability.Flow control610 withinFIG.9 incorporates two flow control elements, a firsthigh flow element901 and a secondlow flow element902. These flow elements may be any flow control device known in the art but, as embodiments,901 and902 are orifices. Orifices provide a low-cost way to regulate airflow rate when in the presence of an applied vacuum, such as that provided bypump627. In the embodiments of900high flow element901 provides an airflow rate necessary to convey the air samples from eachexhaust stream314,315,316,317 tosystem611. This airflow provided by901 (herein purge flow) is set to a relatively high flow rate compared to that provided byflow element902 to enable each of thetubings613,616,619, and622 to be substantially cleared of any previous samples for each step of the sampling sequence. The airflow rate provided by flow element902 (herein sensing flow) need only be a fraction of the purge flow value, as the sensing flow rate need only convey each air sample a short distance (a few inches) fromvalves621,618,615,612 tosensor602. Typically,tubing613,616,619, and622 will be 20 to 50 feet or more in length. The sensing flow rate should also be limited to prevent significant pressure drops withsensor602, which would affect sensor accuracy. As an embodiment,purge flow901 is set to 15 liters per minute andsensing flow902 is set to 2 liters per minute.
• Embodiments ofsystem900 incorporate 3-way valves903 and904 to control the flow rates during each state of the sampling sequence as controlled bylogic605. Like the operation of airsampling system embodiments800,system900 provides sequential air sampling functionality. During normal sampling operation an air sample is conveyed from a location by first placing610 into its purge flow state. During that time common port A ofvalve904 is open to port C, (closed to port B) and common port A ofvalve903 is open to port C. This allows the purge flow rate established byelement901 to be applied to the location being sampled. For example, when the control sequence prompts611 to sample fromexhaust stream317, two-way valve621 will first be opened withvalves612,615, and618 closed. Once the purge flow state has been applied to317 for a predetermined period of time (which can be variable) the flow state will change to sensing mode (low flow), in which the common port A ofvalve904 will be opened to port B and the airflow sample fromtubing622 will be conveyed through904 tosensor602 at the lower flow rate established by902. Like the purge sequence, the sensing sequence is performed for a predetermined period of time. This sensing duration is a function of the response time of thesensor602, which may include a number of sensors. Therefore, the sensing duration will be a function of the slowest acting sensor. In a preferred embodiment, the purge sequence is fixed at 15 seconds in duration and the sensing sequence duration is 15 seconds. It should be clear to those experienced in the art of multipoint air sampling systems that variable purge and sensing times can be applied. For example, in some applications, one or more sensedlocations314,315,316,317 could be farther away than other sensed locations, and that in such applications it can be advantageous to assign a purge time that may be different for each sensed location. Likewise, whensensor602 is composed of a plurality of sensors, it can be advantageous to vary the sensing time based on which sensor is enabled as a location is sampled. Therefore, embodiments of this invention apply to both fixed and variable purge and sensing times. As described byinventive logic500, if contaminant levels that have been sensed bysensor602 by the end of the sampling sequence do not exceed thepredetermined action level515, thensystem900 will continue its sequence by sampling from the next location in the sequence. Alternatively, if the contaminant levels that have been sensed bysensor602 do exceed thepredetermined action level515, thensystem900 will switch into the state of sensorprotective mode509.
• As an embodiment ofsystem900, when the sensorprotective mode state509 has been activated: common port A ofvalve904 will be opened to port C of904, common port A ofvalve903 will be opened to port B of903, two-way valves621,618,615, and612 to each monitored location will be open. This will isolate thesensor602 from the contaminant source (314,315,316, or317) and, in one embodiment,place sensor602 under the full vacuum of627, which acts to evacuate and desorb contaminants that may have setup within602, with ensuressensor602 accuracy and reliability. As an alternate embodiment, which also ensures sensor accuracy and reliability, when in sensorprotective mode509, optional two-way valve905 will open to the atmosphere or ambient air906 to enable fresh air to dilute contaminants withinsensor602, as ambient air flows through905 intosensor602, throughflow element902 and then out tovacuum pump627. Ambient air906 may include any source of clean air, including the air surrounding thesystem900. For example,906 may be air in a mechanical space, outdoor air, or other clean air source. While900 is in the sensorprotective mode state509, the positions ofvalves903 and904, along with open two-way valves621,618,615, and612 creates a path for air to flow from ambient air906, throughvalve903, throughvalve904 and throughvalves621,618,615, and612, to airflowstreams314,315,316, and317. This directional flow from ambient air906 to314,315,316, and317 is due to the negative pressure of the exhaust air caused by theexhaust fans309,310,311. This provides a flushing function totubing613,616,619, and622 that is advantageous as this action removes adsorbed compounds which setup in the tubing that can over time affect the accuracy of the contaminant readings performed bysensor602.
• FIG.10 further illustrates example sampling operation ofsystem900. Shown inFIG.10 are eight states of the valves of611 during normal sampling mode, assuming that900 is configured to monitor four exhaust duct locations (317,316,315,314). However,system900 may be extended to operate with any number of exhaust duct locations.FIG.10 also shows a ninth state which illustrates the valve logic of sensor protection mode forsystem900. Each state of900 is assigned a state number for reference purposes. Based on theconfiguration900 shown,system900 begins samplingair stream317 starting withstate1001, in which900 is placed into purge mode. InFIG.10, the state of each valve or valve port pair (A/B, A/C) is signified as “closed” or “open”. In each case, “closed” signifies the airflow path is blocked and “open” signifies the airflow path is opened. Instate1001, air will flow throughvalve904 from port A to port C and then continue to flow from port A to port C throughvalve903 and throughhigh flow element901. After a predetermined period, such as but not limited to 15 seconds,tubing622 will be adequately purged and the state ofsystem900 will change tostate1002, in whichsystem900 is in sensing mode, as the air sample which was conveyed from317 toopen valve621 is diverted throughvalve904 by opening port A ofvalve904 to port B of904, thus enabling the air sample to flow throughsensor element602 at a flow rate determined byflow element902. At the end of thestate1002,system900 evaluates if the sensed concentration of the sample from317 exceeds thepredetermined action level515. Ifaction level515 is exceeded, thensystem900 will be placed in the sensorprotective mode state509 andfan setback517 signal will be set to the false state, thus disabling fan setback. If the sample from317 does not exceedaction level515,system900 will begin to acquire an air sample in purge mode from the next location viastate1003. This process continues indefinitely until a sensed condition that exceedsaction level515 is encountered, or any number of override conditions (703,704,705) are encountered.
• FIG.10 illustrates a scenario where high contaminant levels (contaminant levels which exceed action level515) are detected instate1008, while samplinglocation314. As shown inFIG.10, said high contaminant levels will causesystem900 to switch intostate1009, which is the sensor protective mode state with the logic which was described above. When instate1009, the sampling operation of900 will be interrupted for a period equal tosequence delay503 andsensor602 will be protected from further exposure to contaminant levels.
• As has been described, thefan setback signal517 derived frominventive logic500 is acted upon by the fan controls orBAS407 to lower the total airflow throughfans309,310, and311 when airflow streams314,315,316,317 are relatively free of contaminants. In systems where bypass air is present, this would be accomplished by reducingbypass air301 until a predetermined minimum exit velocity of312 discharge air is achieved. The end result is a beneficial reduction offan309,310,311 energy consumption and therefore energy cost. For exhaust fan systems which do not incorporatebypass air301 the airflow throughfans309,310,311 is a function only of total system exhaust221, which is determined by the total exhaust flows from each lab or room zone served by the exhaust fan system. Therefore, for systems such as this that do not havebypass301,exhaust fan203 energy reduction cannot be achieved without lowering laboratory flows. Also, as prior art, when laboratory airflow rates or ACH values are reduced in order to save heating and cooling energy,supply fan202 energy, and potentially exhaustfan203 energy, the amount of energy savings that can be achieved is often limited by the amount by which the ECM can reduce theexhaust fan203 exit velocity while ensuringsafe exhaust fan203 operation under all conditions. Where possible, lab airflow reduction ECMs are accomplished by specifying a lower minimum ACH value for each lab than was specified in the original design. If the existingfan203 does not incorporate abypass301 and203 was originally sized to just deliver a minimum acceptable exit velocity at the minimum design exhaust CFM of the labs (forexample exhaust204,205,207, and211), then a lab flow reduction ECM will not be possible, as it would result in unsafe exhaust fan operation during some operating periods where the system exhaust221 is contaminated. As an embodiment of this invention,fan setback signal517 is used to actively enable flow reductions and energy savings infan203 by monitoring exhaust221 usingsystem600 and reducing lab air change rates (ACH) when the exhaust is relatively clean. This may be accomplished by interfacingfan setback signal517 to the BAS or laboratory controls in order to activate a clean exhaust minimum lab ACH value when exhaust221 is determined bylogic500 to be relatively free of contaminants (concentrations sensed bysensor602 are lower than the action level515).FIG.11 illustrates embodiments of usingactive sensing logic500 to reducefan203 energy use via a clean exhaust minimum ACH value. The control logic (clean exhaust minimum ACH logic)1100 incorporates alogic stage1101 that determines if thefan setback signal517 is True or False. If517 is True, the logic proceeds tologic stage1102 which activates a reduced lab minimum CFM to one or multiple labs served byexhaust fan203.Logic1102 is then followed bylogic1101, which again evaluatesfan setback signal517 and if True, the cycle continues throughlogic1102. If inlogic stage1101, it is determined thatfan setback signal517 is False, thenlogic1100 will proceed tologic stage1103, which sets the minimum CFM values in the one or more labs served byexhaust fan203 to their original design minimum value. As prior art, lab minimum CFM is often controlled based on the supply air that is provided to the lab. For example, lab minimum CFM may be determined as the minimum ofsupply air206,209,210,212. However, lab minimum CFM may also be determined as the minimum of the exhaust air that is removed from each lab. For example, it may be the minimum ofexhausts204,205,207,208,211. As embodiments of this invention, the lab minimum CFM implemented bylogic1102 and1103 may be established in terms of either the exhaust or the supply air to each lab. As an embodiment, an instance oflogic1100 is created in software for each lab whose minimum ventilation is to be adjusted in order to effectively lower the total flow and fan energy offan203 when it is determined bylogic500 that thefan203 may be set back. As an embodiment,logic1100 may be implemented in the BAS. As an alternate embodiment,logic1100 may be implemented within a multipoint air sampling system. As an embodiment,1100 may be implemented as a module withinlogic500.
• FIG.12 shows anexemplary computer1200 that can perform at least part of the processing described herein. Thecomputer1200 includes aprocessor1202, avolatile memory1204, a non-volatile memory1206 (e.g., hard disk), anoutput device1207 and a graphical user interface (GUI)1208 (e.g., a mouse, a keyboard, a display, for example). Thenon-volatile memory1206stores computer instructions1212, anoperating system1216 anddata1218. In one example, thecomputer instructions1212 are executed by theprocessor1202 out ofvolatile memory1204. In one embodiment, anarticle1220 comprises non-transitory computer-readable instructions.
• Processing may be implemented in hardware, software, or a combination of the two. Processing may be implemented in computer programs executed on programmable computers/machines that each includes a processor, a storage medium or other article of manufacture that is readable by the processor (including volatile and non-volatile memory and/or storage elements), at least one input device, and one or more output devices. Program code may be applied to data entered using an input device to perform processing and to generate output information.
• Processing may be performed by one or more programmable processors executing one or more computer programs to perform the functions of the system. All or part of the system may be implemented as, special purpose logic circuitry (e.g., an FPGA (field programmable gate array) and/or an ASIC (application-specific integrated circuit)).
• Having described exemplary embodiments of the invention, it will now become apparent to one of ordinary skill in the art that other embodiments incorporating their concepts may also be used. The embodiments contained herein should not be limited to disclosed embodiments but rather should be limited only by the spirit and scope of the appended claims. All publications and references cited herein are expressly incorporated herein by reference in their entirety.
• Elements of different embodiments described herein may be combined to form other embodiments not specifically set forth above. Various elements, which are described in the context of a single embodiment, may also be provided separately or in any suitable subcombination. Other embodiments not specifically described herein are also within the scope of the following claims.

Claims (20)

What is claimed is:

1. An exhaust demand control system for measuring one or more contaminants at one or more exhaust locations within one or more exhaust ducts or plenums served by an exhaust fan system having one or more exhaust fans, comprising systems configured to:

sense the one or more contaminants within the one or more exhaust locations using a multipoint air sampling system having one or more sensors and flow control elements, which include a high flow element to provide a purge flow and a low flow element to provide a sensing flow;

compare contaminant concentration measurements from the one or more exhaust locations against an action level to create a fan setback signal which controls a setback state of the exhaust fan system;

prevent exhaust fan system instability by controlling operation of the one or more exhaust fans according to a sequence delay provided by control logic of a control system where the sequence delay is enabled when the contaminant concentration measurements from the one or more exhaust locations exceeds the action level, wherein the sequence delay in part determines a minimum time that the fan setback signal is disabled; and

ensure sensor accuracy and reliability by implementing a sensor protective mode for a period of time designated by the sequence delay, wherein the sensor protective mode is enabled when the contaminant concentration measurements from the one or more exhaust locations exceeds the action level, wherein the sensor protective mode includes the combined steps of: discontinuing an air sampling process performed by the multipoint air sampling system, isolating the one or more sensors from the contaminant source, and enabling a tubing flushing function which uses a negative pressure of the one or more exhaust locations to flow ambient air through each tube which connects between the multipoint air sampling system and the one or more exhaust locations via a flow path to ambient that is enabled within the multipoint air sampling system when the sensor protective mode is enabled.

2. The exhaust demand control system ofclaim 1, wherein the exhaust fan system includes one or more high plume fans.

3. The exhaust demand control system ofclaim 1, wherein at least a portion of the one or more exhaust ducts includes one or more exhaust risers.

4. The exhaust demand control system ofclaim 1 which, as an added measure to ensure sensor accuracy and reliability while isolating the one or more sensors from the contaminant source, places the one or more sensors under vacuum to evacuate and desorb contaminants that the one or more sensors have been exposed to.

5. The exhaust demand control system ofclaim 1, wherein the systems are further configured to: while isolating the one or more sensors from the contaminant source, introduce a source of clean air to the one or more sensors to dilute contaminants within the one or more sensors.

6. The exhaust demand control system ofclaim 1, wherein the high and low flow elements of the multipoint air sampling system individually comprise an orifice which each regulate airflow rate using an applied vacuum from a vacuum pump.

7. The exhaust demand control system ofclaim 1, wherein the multipoint air sampling system comprises a networked air sampling system.

8. The exhaust demand control system ofclaim 1, that further incorporates a setback override function based on one or more error conditions to limit when the exhaust fan system is set back.

9. The exhaust demand control system ofclaim 8, wherein the one or more error conditions includes a failed vacuum pump.

10. The exhaust demand control system ofclaim 1, wherein the one or more sensors comprises a Photoionization Detector (PID) sensor.

11. The exhaust demand control system ofclaim 1, wherein the sequence delay is adaptive.

12. The exhaust demand control system ofclaim 1, wherein the exhaust demand control function incorporates clean exhaust minimum ACH logic.

13. The exhaust demand control system ofclaim 1, wherein the multipoint air sampling system is contained within a single enclosure.

14. A method of performing exhaust demand control using a system for measuring one or more contaminants at one or more exhaust locations within one or more exhaust ducts or plenums served by an exhaust fan system having one or more exhaust fans, comprising:

sensing the one or more contaminants within the one or more exhaust locations using a multipoint air sampling system having one or more sensors and two flow control elements which include a high flow element to provide a purge flow and a low flow element to provide a sensing flow;

comparing contaminant concentration measurements from the one or more exhaust locations against an action level to create a fan setback signal which controls the setback state of the exhaust fan system;

preventing exhaust fan system instability by controlling the operation of the one or more exhaust fans according to a sequence delay provided by control logic of the control system where the sequence delay is enabled when the contaminant concentration measurements from the one or more exhaust locations exceeds the action level, wherein the sequence delay in part determines a minimum time that the fan setback signal is disabled; and

ensuring sensor accuracy and reliability by implementing a sensor protective mode for a period of time designated by the sequence delay, which sensor protective mode is enabled when the contaminant concentration measurements from the one or more exhaust locations exceeds the action level; sensor protective mode at least consisting of the combined steps of: discontinuing the air sampling process performed by the multipoint air sampling system, isolating the one or more sensors from the contaminant source, and enabling a tubing flushing function which uses the negative pressure of the one or more exhaust locations to flow ambient air through each tube which connects between the multipoint air sampling system and the one or more exhaust locations via a flow path to ambient that is enabled within the multipoint air sampling system when sensor protective mode is enabled.

15. The method ofclaim 14, wherein the exhaust fan system includes one or more high plume fans.

16. The method ofclaim 14, wherein at least a portion of the one or more exhaust ducts includes one or more exhaust risers.

17. The method ofclaim 14 wherein, as an added measure to ensure sensor accuracy and reliability while isolating the one or more sensors from the contaminant source, places the one or more sensors under vacuum to evacuate and desorb contaminants.

18. The method ofclaim 14 wherein, as an added measure to ensure sensor accuracy and reliability while isolating the one or more sensors from the contaminant source, introduces a source of clean air to the one or more sensors to dilute contaminants within the one or more sensors.

19. The method ofclaim 14, wherein the multipoint air sampling system comprises a networked air sampling system.

20. The method ofclaim 14, wherein the one or more sensors comprises a Photoionization Detector (PID) sensor.

Images